{ "prompt": "A playful orange tabby cat starts in an ordinary cozy living room in a normal house, afternoon light, sofa and rug. The cat crouches, jumps, and does one clean athletic backflip in slow motion. As the backflip completes there is a sharp cinematic cut: the cat lands perfectly on a glowing neon disco dance floor wearing oversized black sunglasses. Mirror ball reflections, colorful lights, joyful party energy, stylish and funny, clear before-and-after transformation.", "model_path": "/models/minimax_h3_fl2va_pruned_nvfp4.safetensors", "width": 1344, "height": 768, "frames": 124, "steps": 12, "sampler_step": 1, "seed": 440420, "text_tokens": 100, "block_index": 24, "attention": "sage2", "warmup": 20, "iterations": 50, "hidden_shape": [ 37810, 5376 ], "output_shape": [ 37810, 5376 ], "segments": [ [ 0, 100, 1 ], [ 100, 514, 2 ], [ 514, 37810, 0 ] ], "timings": { "norm1": { "count": 50, "mean_s": 0.0035359218600569875, "p50_s": 0.0035142429996994906, "p90_s": 0.003590562999852409, "p95_s": 0.0036120837999078503, "p99_s": 0.003746741560207738, "min_s": 0.0034944029994221637, "max_s": 0.003793122999923071 }, "modulate_msa": { "count": 50, "mean_s": 0.010587173820022144, "p50_s": 0.01043039999967732, "p90_s": 0.010989329001313308, "p95_s": 0.010999922849896393, "p99_s": 0.011023483050266805, "min_s": 0.010271736000504461, "max_s": 0.011034584000299219 }, "linear.attn_qkv_proj.flatten_contiguous": { "count": 50, "mean_s": 4.358399964985438e-06, "p50_s": 4.255999556335155e-06, "p90_s": 4.454400368558709e-06, "p95_s": 4.644000000553205e-06, "p99_s": 6.431039619201322e-06, "min_s": 4.079998689121567e-06, "max_s": 7.951999577926472e-06 }, "linear.attn_qkv_proj.pre_quant_scale": { "count": 50, "mean_s": 0.0, "p50_s": 0.0, "p90_s": 0.0, "p95_s": 0.0, "p99_s": 0.0, "min_s": 0.0, "max_s": 0.0 }, "linear.attn_qkv_proj.packed_weight_wrapper": { "count": 50, "mean_s": 8.606720220996068e-06, "p50_s": 8.119999620248564e-06, "p90_s": 8.673600859765429e-06, "p95_s": 9.055200825969221e-06, "p99_s": 1.9636799952422696e-05, "min_s": 7.58399983169511e-06, "max_s": 2.927999958046712e-05 }, "linear.attn_qkv_proj.bias_cast": { "count": 50, "mean_s": 3.511040013108868e-06, "p50_s": 3.4720005714916624e-06, "p90_s": 3.604800258472096e-06, "p95_s": 3.7135997445147947e-06, "p99_s": 3.989119686593766e-06, "min_s": 3.3760006772354245e-06, "max_s": 4.239998816046864e-06 }, "linear.attn_qkv_proj.activation_scale": { "count": 50, "mean_s": 0.0017170804400302587, "p50_s": 0.0017020014993249788, "p90_s": 0.0017806290994485607, "p95_s": 0.0017970364000575501, "p99_s": 0.0018388336505449842, "min_s": 0.0016812330013635801, "max_s": 0.0018447210004524095 }, "linear.attn_qkv_proj.scale_to_device": { "count": 50, "mean_s": 1.0970560033456423e-05, "p50_s": 9.424000381841324e-06, "p90_s": 1.3892800961912145e-05, "p95_s": 1.8935199113911944e-05, "p99_s": 2.951103981104094e-05, "min_s": 8.319999324157834e-06, "max_s": 3.7695999708375894e-05 }, "linear.attn_qkv_proj.activation_quant_pack": { "count": 50, "mean_s": 0.0021769854400554324, "p50_s": 0.0021467699989443645, "p90_s": 0.0022839940002086223, "p95_s": 0.0023105996001504536, "p99_s": 0.002333132719668356, "min_s": 0.002122096999300993, "max_s": 0.0023364019998552976 }, "linear.attn_qkv_proj.activation_quant_wrap": { "count": 50, "mean_s": 6.238080022740177e-06, "p50_s": 6.159999429655727e-06, "p90_s": 6.5759997596615e-06, "p95_s": 6.858400229248218e-06, "p99_s": 7.811040286469504e-06, "min_s": 5.727999450755306e-06, "max_s": 8.352000804734416e-06 }, "linear.attn_qkv_proj.gemm": { "count": 50, "mean_s": 0.026236184619883716, "p50_s": 0.026313476000723313, "p90_s": 0.0267578761002369, "p95_s": 0.0267910313997163, "p99_s": 0.026870087120514655, "min_s": 0.025206532998709008, "max_s": 0.02692462999948475 }, "linear.attn_qkv_proj.slice_reshape": { "count": 50, "mean_s": 5.6448200120939875e-06, "p50_s": 5.576001058216207e-06, "p90_s": 5.824000254506245e-06, "p95_s": 5.882399636902846e-06, "p99_s": 7.303040310944193e-06, "min_s": 5.296000381349586e-06, "max_s": 8.431999958702363e-06 }, "attn_qkv_proj": { "count": 50, "mean_s": 0.030210273239936213, "p50_s": 0.03035905599972466, "p90_s": 0.030703277699649333, "p95_s": 0.03072695820019362, "p99_s": 0.030874537349391176, "min_s": 0.029219621999800438, "max_s": 0.03091569000025629 }, "attn_qkv_split_view": { "count": 50, "mean_s": 1.1833939897769596e-05, "p50_s": 1.1328000255161896e-05, "p90_s": 1.3222398774814793e-05, "p95_s": 1.4168800498737254e-05, "p99_s": 1.9776160061155663e-05, "min_s": 1.046399847837165e-05, "max_s": 2.4095999833662063e-05 }, "attn_qk_rms_rope": { "count": 50, "mean_s": 0.012088846719998401, "p50_s": 0.012047434000123758, "p90_s": 0.012417796300906048, "p95_s": 0.012425689800329565, "p99_s": 0.01257420811052725, "min_s": 0.011746297001081984, "max_s": 0.012706587000138825 }, "attention_kernel": { "count": 50, "mean_s": 0.26502367770001, "p50_s": 0.265111950499886, "p90_s": 0.2670563616989966, "p95_s": 0.26712134680019517, "p99_s": 0.2677792321504239, "min_s": 0.262332772001173, "max_s": 0.26781361299981654 }, "attn_output_reshape": { "count": 50, "mean_s": 4.21248005295638e-06, "p50_s": 4.160000571573619e-06, "p90_s": 4.355199598649051e-06, "p95_s": 4.7128000915108714e-06, "p99_s": 4.971359285264043e-06, "min_s": 3.935998392989859e-06, "max_s": 5.135998435434885e-06 }, "linear.attn_out_proj.flatten_contiguous": { "count": 50, "mean_s": 4.259839988662861e-06, "p50_s": 4.175999492872506e-06, "p90_s": 4.68319976789644e-06, "p95_s": 4.8599995352560654e-06, "p99_s": 5.284480539557989e-06, "min_s": 3.952000042772852e-06, "max_s": 5.504000000655651e-06 }, "linear.attn_out_proj.pre_quant_scale": { "count": 50, "mean_s": 0.0, "p50_s": 0.0, "p90_s": 0.0, "p95_s": 0.0, "p99_s": 0.0, "min_s": 0.0, "max_s": 0.0 }, "linear.attn_out_proj.packed_weight_wrapper": { "count": 50, "mean_s": 9.12031999177998e-06, "p50_s": 8.799999704933725e-06, "p90_s": 9.904000762617216e-06, "p95_s": 1.0727198878157648e-05, "p99_s": 1.329311962763313e-05, "min_s": 8.143999366438948e-06, "max_s": 1.3935999959358014e-05 }, "linear.attn_out_proj.bias_cast": { "count": 50, "mean_s": 3.559359975042753e-06, "p50_s": 3.527999979269225e-06, "p90_s": 3.7664001865778117e-06, "p95_s": 3.832799484371207e-06, "p99_s": 3.96991988964146e-06, "min_s": 3.375998858246021e-06, "max_s": 4.064000677317381e-06 }, "linear.attn_out_proj.activation_scale": { "count": 50, "mean_s": 0.0022237825599950157, "p50_s": 0.0022224574986466905, "p90_s": 0.0022319219999189953, "p95_s": 0.0022335843996188487, "p99_s": 0.0022542973603594876, "min_s": 0.0022130420002213214, "max_s": 0.0022664179996354505 }, "linear.attn_out_proj.scale_to_device": { "count": 50, "mean_s": 1.0166740030399524e-05, "p50_s": 9.976000001188368e-06, "p90_s": 1.0788800500449724e-05, "p95_s": 1.1143999745399921e-05, "p99_s": 1.3063200240139847e-05, "min_s": 8.991999493446201e-06, "max_s": 1.3807999494019896e-05 }, "linear.attn_out_proj.activation_quant_pack": { "count": 50, "mean_s": 0.0028620550799314513, "p50_s": 0.0028830265000578947, "p90_s": 0.0028994244999921647, "p95_s": 0.002903895250892674, "p99_s": 0.002973585690124309, "min_s": 0.0028147700013505528, "max_s": 0.0030201940007827943 }, "linear.attn_out_proj.activation_quant_wrap": { "count": 50, "mean_s": 6.35775981209008e-06, "p50_s": 6.296000719885342e-06, "p90_s": 6.663998829026241e-06, "p95_s": 6.848000339232385e-06, "p99_s": 7.630399486515668e-06, "min_s": 5.935999070061371e-06, "max_s": 8.335999154951423e-06 }, "linear.attn_out_proj.gemm": { "count": 50, "mean_s": 0.008039239360005012, "p50_s": 0.008028734499930579, "p90_s": 0.008143914901120297, "p95_s": 0.008157982450484269, "p99_s": 0.00827479994910391, "min_s": 0.007810614000845817, "max_s": 0.008372870999664883 }, "linear.attn_out_proj.slice_reshape": { "count": 50, "mean_s": 5.5884800531202926e-06, "p50_s": 5.583999154623598e-06, "p90_s": 5.7791990911937315e-06, "p95_s": 5.81680033064913e-06, "p99_s": 5.995040301058907e-06, "min_s": 5.312000212143175e-06, "max_s": 6.144000508356839e-06 }, "attn_out_proj": { "count": 50, "mean_s": 0.013204702960210852, "p50_s": 0.013201394999668992, "p90_s": 0.013340711800265127, "p95_s": 0.01335459340034504, "p99_s": 0.01355505611976696, "min_s": 0.012927738000144018, "max_s": 0.013737978999415645 }, "gate_msa": { "count": 50, "mean_s": 0.008690837319918501, "p50_s": 0.008646599499115837, "p90_s": 0.008749174999866226, "p95_s": 0.009079649749855888, "p99_s": 0.009193422359730903, "min_s": 0.008623383000667673, "max_s": 0.00926199099922087 }, "norm2": { "count": 50, "mean_s": 0.003568036460201256, "p50_s": 0.003531907000251522, "p90_s": 0.0036303837016021133, "p95_s": 0.003743043001395563, "p99_s": 0.0038393764505599394, "min_s": 0.003515746999255498, "max_s": 0.0038490269998874282 }, "modulate_mlp": { "count": 50, "mean_s": 0.010634301519858128, "p50_s": 0.010705913000492728, "p90_s": 0.010969404899333313, "p95_s": 0.01097855734933546, "p99_s": 0.01102472357910301, "min_s": 0.010306168000170146, "max_s": 0.011047977999623981 }, "linear.mlp_fc1.flatten_contiguous": { "count": 50, "mean_s": 4.33599987445632e-06, "p50_s": 4.303999958210625e-06, "p90_s": 4.4959997467231005e-06, "p95_s": 4.527999408310279e-06, "p99_s": 5.126080432091839e-06, "min_s": 4.064000677317381e-06, "max_s": 5.424000846687704e-06 }, "linear.mlp_fc1.pre_quant_scale": { "count": 50, "mean_s": 0.0, "p50_s": 0.0, "p90_s": 0.0, "p95_s": 0.0, "p99_s": 0.0, "min_s": 0.0, "max_s": 0.0 }, "linear.mlp_fc1.packed_weight_wrapper": { "count": 50, "mean_s": 8.286079973913729e-06, "p50_s": 8.199999683711212e-06, "p90_s": 8.604798495071009e-06, "p95_s": 8.881600660970434e-06, "p99_s": 1.04201599060616e-05, "min_s": 7.728000127826817e-06, "max_s": 1.062400042428635e-05 }, "linear.mlp_fc1.bias_cast": { "count": 50, "mean_s": 3.53920000634389e-06, "p50_s": 3.48799949279055e-06, "p90_s": 3.633599772001617e-06, "p95_s": 3.6976009141653774e-06, "p99_s": 4.592639797920124e-06, "min_s": 3.3599990274524316e-06, "max_s": 5.4079991969047114e-06 }, "linear.mlp_fc1.activation_scale": { "count": 50, "mean_s": 0.0017303090999121196, "p50_s": 0.0017195689997606678, "p90_s": 0.0018050011001832901, "p95_s": 0.0018161890008741464, "p99_s": 0.0018366168388274672, "min_s": 0.0016687049992469838, "max_s": 0.0018491529990569688 }, "linear.mlp_fc1.scale_to_device": { "count": 50, "mean_s": 1.0146900021936745e-05, "p50_s": 9.456000043428503e-06, "p90_s": 1.0617600673867857e-05, "p95_s": 1.2792001143679946e-05, "p99_s": 2.6764480571728167e-05, "min_s": 8.479999451083131e-06, "max_s": 3.44320014846744e-05 }, "linear.mlp_fc1.activation_quant_pack": { "count": 50, "mean_s": 0.0022118504599347943, "p50_s": 0.0022029219999240013, "p90_s": 0.0022765955998693245, "p95_s": 0.0023190750502180887, "p99_s": 0.00235583480061905, "min_s": 0.0021436339993670117, "max_s": 0.0023676340006204555 }, "linear.mlp_fc1.activation_quant_wrap": { "count": 50, "mean_s": 6.254719919525087e-06, "p50_s": 6.223999662324786e-06, "p90_s": 6.531200051540508e-06, "p95_s": 6.748800569766899e-06, "p99_s": 7.058879964461084e-06, "min_s": 5.840000085299835e-06, "max_s": 7.200000254670158e-06 }, "linear.mlp_fc1.gemm": { "count": 50, "mean_s": 0.03515858558006585, "p50_s": 0.03511626750059804, "p90_s": 0.03549363759975677, "p95_s": 0.03557769874960286, "p99_s": 0.03572802138010957, "min_s": 0.03414487400004873, "max_s": 0.03578301800007466 }, "linear.mlp_fc1.slice_reshape": { "count": 50, "mean_s": 5.572800073423423e-06, "p50_s": 5.551999493036419e-06, "p90_s": 5.761599277320784e-06, "p95_s": 5.841600068379194e-06, "p99_s": 6.067840458854333e-06, "min_s": 5.343999873730354e-06, "max_s": 6.2560011429013684e-06 }, "mlp_fc1": { "count": 50, "mean_s": 0.03918287416006933, "p50_s": 0.03917559199999232, "p90_s": 0.039542551699923933, "p95_s": 0.039647989400327786, "p99_s": 0.039722929230501906, "min_s": 0.038156013000843814, "max_s": 0.03972972700103128 }, "mlp_swiglu": { "count": 50, "mean_s": 0.024074891440068313, "p50_s": 0.024049810999713372, "p90_s": 0.024356551100390787, "p95_s": 0.024452045749512763, "p99_s": 0.024497954870494137, "min_s": 0.023798129999704543, "max_s": 0.024512820000381907 }, "linear.mlp_fc2.flatten_contiguous": { "count": 50, "mean_s": 4.211519808450248e-06, "p50_s": 4.167999577475712e-06, "p90_s": 4.403200546221342e-06, "p95_s": 4.511200586421182e-06, "p99_s": 4.908320879621896e-06, "min_s": 3.9360002119792625e-06, "max_s": 5.1200004236306995e-06 }, "linear.mlp_fc2.pre_quant_scale": { "count": 50, "mean_s": 0.0, "p50_s": 0.0, "p90_s": 0.0, "p95_s": 0.0, "p99_s": 0.0, "min_s": 0.0, "max_s": 0.0 }, "linear.mlp_fc2.packed_weight_wrapper": { "count": 50, "mean_s": 7.91585993283661e-06, "p50_s": 7.840000762371346e-06, "p90_s": 8.257599984062835e-06, "p95_s": 8.703200273885157e-06, "p99_s": 9.320799654233269e-06, "min_s": 7.296001058421098e-06, "max_s": 9.359999239677563e-06 }, "linear.mlp_fc2.bias_cast": { "count": 50, "mean_s": 3.5664000824908725e-06, "p50_s": 3.5360008041607216e-06, "p90_s": 3.684801049530506e-06, "p95_s": 3.763200220419094e-06, "p99_s": 3.9768007081875104e-06, "min_s": 3.3760006772354245e-06, "max_s": 4.016001184936613e-06 }, "linear.mlp_fc2.activation_scale": { "count": 50, "mean_s": 0.004592374580024625, "p50_s": 0.004559275500469084, "p90_s": 0.004735260000234121, "p95_s": 0.004762049599139573, "p99_s": 0.004766211350379308, "min_s": 0.004506786999627366, "max_s": 0.00476659600099083 }, "linear.mlp_fc2.scale_to_device": { "count": 50, "mean_s": 1.0288320081599523e-05, "p50_s": 9.424000381841324e-06, "p90_s": 1.2665600115724375e-05, "p95_s": 1.4833600562269567e-05, "p99_s": 2.0252960439393047e-05, "min_s": 8.70400072017219e-06, "max_s": 2.4416000087512657e-05 }, "linear.mlp_fc2.activation_quant_pack": { "count": 50, "mean_s": 0.005908733240030415, "p50_s": 0.005951356499281246, "p90_s": 0.006028886500280351, "p95_s": 0.0060370161999344415, "p99_s": 0.0060677086695250185, "min_s": 0.00564893199953076, "max_s": 0.0060685009993903805 }, "linear.mlp_fc2.activation_quant_wrap": { "count": 50, "mean_s": 6.152319983812049e-06, "p50_s": 6.039999789209105e-06, "p90_s": 6.579199725820218e-06, "p95_s": 6.739199670846574e-06, "p99_s": 7.526239696744594e-06, "min_s": 5.711999619961716e-06, "max_s": 8.224000339396298e-06 }, "linear.mlp_fc2.gemm": { "count": 50, "mean_s": 0.05473969570000918, "p50_s": 0.054685164999682456, "p90_s": 0.05519963150054536, "p95_s": 0.05526944609919156, "p99_s": 0.05548157012059164, "min_s": 0.05419282500042755, "max_s": 0.05548338900007366 }, "linear.mlp_fc2.slice_reshape": { "count": 50, "mean_s": 5.6211200353573075e-06, "p50_s": 5.527999746846035e-06, "p90_s": 5.95199890085496e-06, "p95_s": 6.123199909779941e-06, "p99_s": 6.801280833315104e-06, "min_s": 5.2320010581752285e-06, "max_s": 6.8640001700259745e-06 }, "mlp_fc2": { "count": 50, "mean_s": 0.06531846147994656, "p50_s": 0.06529986749956151, "p90_s": 0.06578957450019515, "p95_s": 0.06586715780076702, "p99_s": 0.06601896072916133, "min_s": 0.0645643679999921, "max_s": 0.06614333399920724 }, "gate_mlp": { "count": 50, "mean_s": 0.008981906899825844, "p50_s": 0.0086321990002034, "p90_s": 0.009006409600442567, "p95_s": 0.009076948600159084, "p99_s": 0.01610825177995136, "min_s": 0.008592614998633508, "max_s": 0.022754208999685943 }, "block_total": { "count": 50, "mean_s": 0.4951837136800532, "p50_s": 0.4949583219995475, "p90_s": 0.49724229510084117, "p95_s": 0.497455911299312, "p99_s": 0.502053749699644, "min_s": 0.49234532799891895, "max_s": 0.5055920739996509 } }, "module_forward": { "count": 50, "mean_s": 0.42752286658007504, "p50_s": 0.42741008050052187, "p90_s": 0.42897752329972716, "p95_s": 0.42947807479968103, "p99_s": 0.43056982682959644, "min_s": 0.4248241990007955, "max_s": 0.4311483860001317 }, "module_forward_checksum": -15117459456.0, "fused_elementwise": true, "profiler_summary": { "profiled_iterations": 1, "runtime_kernel_launches": 0, "runtime_kernel_launches_per_block": 0.0, "positive_self_device_allocated_bytes": 8009579008 }, "profiler_top_events": [ { "key": "aten::rms_norm", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 1546.5930000000008, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": -303104, "input_shapes": "[[37810, 5376], [], [5376], []]" }, { "key": "aten::_fused_rms_norm", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 1535.761, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 5376], [], [5376], []]" }, { "key": "aten::linear", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 1039.4409999999993, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 5376], [21504, 5376], []]" }, { "key": "comfy_kitchen::scaled_mm_nvfp4", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 822.1930000000002, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": -512, "input_shapes": "[[37824, 2688], [21504, 2688], [], [], [37888, 336], [21504, 336], [], [], []]" }, { "key": "comfy_kitchen::rms_rope_split_half_", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 339.29699999999957, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[1, 37810, 56, 128], [1, 37810, 56, 128], [1, 37810, 1, 48, 2, 2], [128], [128], [], []]" }, { "key": "aten::contiguous", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 195.40800000000013, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[3, 5376], []]" }, { "key": "aten::clone", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 185.59999999999877, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[3, 5376], []]" }, { "key": "aten::zeros", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 155.15199999999913, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[], [], [], [], []]" }, { "key": "aten::empty", "count": 43, "device_type": "DeviceType.CPU", "cpu_time_total_us": 138.87999999999772, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 8009561088, "input_shapes": "[[], [], [], [], [], []]" }, { "key": "aten::linear", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 136.91200000000026, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 7168], [5376, 7168], []]" }, { "key": "aten::copy_", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 132.60799999999972, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[3, 5376], [3, 5376], []]" }, { "key": "comfy_kitchen::scaled_mm_nvfp4", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 118.55999999999949, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": -512, "input_shapes": "[[37824, 3584], [5376, 3584], [], [], [37888, 448], [5376, 448], [], [], []]" }, { "key": "aten::zero_", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 104.68799999999919, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37888, 336]]" }, { "key": "comfy_kitchen::quantize_nvfp4", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 101.3119999999999, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 7168], [], [], [], []]" }, { "key": "aten::mean", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 95.15200000000004, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 14336, "input_shapes": "[[1, 37810, 56, 128], [], [], []]" }, { "key": "sageattention_sm89::qk_int8_sv_f8_accum_f16_fuse_v_scale_attn_inst_buf", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 94.65599999999995, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[1, 37810, 56, 128], [1, 37810, 56, 128], [1, 128, 56, 37824], [1, 37810, 56, 128], [1, 56, 1184], [1, 56, 591], [1, 56, 128], [], [], [], [], []]" }, { "key": "aten::linear", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 92.51199999999972, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 5376], [28672, 5376], []]" }, { "key": "comfy_kitchen::scaled_mm_nvfp4", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 78.20799999999963, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": -512, "input_shapes": "[[37824, 2688], [28672, 2688], [], [], [37888, 336], [28672, 336], [], [], []]" }, { "key": "aten::mul", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 71.6319999999996, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 2048, "input_shapes": "[[], []]" }, { "key": "aten::_to_copy", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 59.3769999999995, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[], [], [], [], [], [], []]" }, { "key": "aten::to", "count": 9, "device_type": "DeviceType.CPU", "cpu_time_total_us": 46.91300000000001, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[], [], [], [], []]" }, { "key": "aten::fill_", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 44.735999999999876, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37888, 336], []]" }, { "key": "aten::copy_", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 40.20899999999983, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[], [], []]" }, { "key": "aten::split", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 32.384000000000015, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 21504], [], []]" }, { "key": "aten::empty_like", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 26.607999999999493, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[3, 5376], [], [], [], [], []]" }, { "key": "aten::to", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 24.047999999999774, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[], [], [], [], [], []]" }, { "key": "aten::narrow", "count": 3, "device_type": "DeviceType.CPU", "cpu_time_total_us": 18.367999999999483, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 21504], [], [], []]" }, { "key": "aten::zeros_like", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 16.20800000000054, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[1], [], [], [], [], []]" }, { "key": "aten::reshape", "count": 5, "device_type": "DeviceType.CPU", "cpu_time_total_us": 16.112000000000535, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[], []]" }, { "key": "aten::slice", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 16.096000000000004, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37824, 21504], [], [], [], []]" }, { "key": "aten::empty_strided", "count": 3, "device_type": "DeviceType.CPU", "cpu_time_total_us": 11.568000000000211, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 1536, "input_shapes": "[[], [], [], [], [], []]" }, { "key": "aten::zero_", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 10.78399999999965, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[1]]" }, { "key": "aten::fill_", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 9.583999999999833, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[1], []]" }, { "key": "aten::view", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 9.488000000000284, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37888, 336], []]" }, { "key": "aten::zero_", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 8.943999999999505, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37888, 448]]" }, { "key": "aten::squeeze", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 7.632000000000517, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[1, 1, 56, 128], []]" }, { "key": "aten::view", "count": 5, "device_type": "DeviceType.CPU", "cpu_time_total_us": 7.376000000002023, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[], []]" }, { "key": "aten::fill_", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 6.943999999999505, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37888, 448], []]" }, { "key": "aten::slice", "count": 3, "device_type": "DeviceType.CPU", "cpu_time_total_us": 6.752000000001317, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 21504], [], [], [], []]" }, { "key": "aten::zero_", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 6.496000000000095, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37888, 896]]" }, { "key": "aten::to", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 5.280000000000399, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[5376], [], [], [], []]" }, { "key": "aten::fill_", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 5.168000000000575, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37888, 896], []]" }, { "key": "aten::as_strided", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 5.023999999999887, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37824, 21504], [], [], []]" }, { "key": "aten::view", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 4.879999999999882, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810], []]" }, { "key": "aten::slice", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 3.8400000000001455, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37824, 5376], [], [], [], []]" }, { "key": "aten::view", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 3.7440000000005966, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 7168], []]" }, { "key": "aten::empty_like", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 3.344000000000051, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[1], [], [], [], [], []]" }, { "key": "aten::as_strided", "count": 3, "device_type": "DeviceType.CPU", "cpu_time_total_us": 2.256000000000313, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 21504], [], [], []]" }, { "key": "aten::reshape", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 2.207999999999629, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 7168], []]" }, { "key": "aten::alias", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 2.1279999999997017, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 21504]]" }, { "key": "aten::slice", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 2.0, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37824, 28672], [], [], [], []]" }, { "key": "aten::view", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 1.9520000000002256, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[21504, 336], []]" }, { "key": "aten::as_strided", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 1.88799999999992, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[1, 37810, 56, 128], [], [], []]" }, { "key": "aten::reshape", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 1.8079999999999927, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[1, 37810, 56, 128], []]" }, { "key": "aten::reshape", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 1.199999999999818, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 5376], []]" }, { "key": "aten::as_strided", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.9599999999991269, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37824, 5376], [], [], []]" }, { "key": "aten::alias", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.8640000000004875, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[1, 37810, 56, 128]]" }, { "key": "aten::view", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.7359999999998763, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37888, 448], []]" }, { "key": "aten::as_strided", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.4960000000000946, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37824, 28672], [], [], []]" }, { "key": "aten::view", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.4320000000006985, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[28672, 336], []]" }, { "key": "aten::view", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.41600000000016735, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[1, 37810, 56, 128], []]" }, { "key": "aten::as_strided", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.3999999999996362, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[1, 1, 56, 128], [], [], []]" }, { "key": "aten::view", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.3999999999996362, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[5376, 448], []]" }, { "key": "aten::alias", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.3999999999996362, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 5376]]" }, { "key": "aten::view", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.38400000000001455, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 5376], []]" }, { "key": "aten::alias", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.35199999999986176, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 28672]]" }, { "key": "aten::view", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.3040000000000873, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37888, 896], []]" }, { "key": "[memory]", "count": 38, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.0, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": -8009274368, "input_shapes": "[]" } ] }