{ "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": 3, "iterations": 10, "hidden_shape": [ 37810, 5376 ], "output_shape": [ 37810, 5376 ], "segments": [ [ 0, 100, 1 ], [ 100, 514, 2 ], [ 514, 37810, 0 ] ], "timings": { "norm1": { "count": 10, "mean_s": 0.003536642901599407, "p50_s": 0.0034976669994648546, "p90_s": 0.0035808175161946565, "p95_s": 0.003707458774442784, "p99_s": 0.003808771781041287, "min_s": 0.0034645950072444975, "max_s": 0.0038341000326909125 }, "modulate_msa": { "count": 10, "mean_s": 0.010376643878407777, "p50_s": 0.01049173646606505, "p90_s": 0.010621935367817059, "p95_s": 0.010627220183960163, "p99_s": 0.010631448036874644, "min_s": 0.01000498398207128, "max_s": 0.010632505000103265 }, "linear.attn_qkv_proj.flatten_contiguous": { "count": 10, "mean_s": 4.6912056859582664e-06, "p50_s": 4.264031304046512e-06, "p90_s": 4.956789780408142e-06, "p95_s": 6.742397090420122e-06, "p99_s": 8.170882938429714e-06, "min_s": 4.0800077840685844e-06, "max_s": 8.52800440043211e-06 }, "linear.attn_qkv_proj.pre_quant_scale": { "count": 10, "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": 10, "mean_s": 1.067199045792222e-05, "p50_s": 8.656003046780825e-06, "p90_s": 1.1094362707808606e-05, "p95_s": 2.0699159358628073e-05, "p99_s": 2.838299667928368e-05, "min_s": 8.01600981503725e-06, "max_s": 3.0303956009447575e-05 }, "linear.attn_qkv_proj.bias_cast": { "count": 10, "mean_s": 3.587285755202174e-06, "p50_s": 3.54400253854692e-06, "p90_s": 3.7352961953729388e-06, "p95_s": 3.907622885890305e-06, "p99_s": 4.045484238304198e-06, "min_s": 3.375986125320196e-06, "max_s": 4.079949576407671e-06 }, "linear.attn_qkv_proj.activation_scale": { "count": 10, "mean_s": 0.001664134208112955, "p50_s": 0.0016452169802505523, "p90_s": 0.0017439107294194402, "p95_s": 0.0017465958721004427, "p99_s": 0.0017487439862452446, "min_s": 0.0016137940110638738, "max_s": 0.0017492810147814453 }, "linear.attn_qkv_proj.scale_to_device": { "count": 10, "mean_s": 1.0867195669561625e-05, "p50_s": 9.648007107898593e-06, "p90_s": 1.2599985348060724e-05, "p95_s": 1.580399402882903e-05, "p99_s": 1.836720097344369e-05, "min_s": 9.232026059180498e-06, "max_s": 1.900800270959735e-05 }, "linear.attn_qkv_proj.activation_quant_pack": { "count": 10, "mean_s": 0.0021940851060207934, "p50_s": 0.002150969987269491, "p90_s": 0.00229719239869155, "p95_s": 0.0023544612020486968, "p99_s": 0.002400276244734414, "min_s": 0.0021200490300543606, "max_s": 0.0024117300054058433 }, "linear.attn_qkv_proj.activation_quant_wrap": { "count": 10, "mean_s": 6.774382200092077e-06, "p50_s": 6.360001862049103e-06, "p90_s": 7.612811168655753e-06, "p95_s": 8.750418783165512e-06, "p99_s": 9.660504874773325e-06, "min_s": 6.175949238240719e-06, "max_s": 9.888026397675276e-06 }, "linear.attn_qkv_proj.gemm": { "count": 10, "mean_s": 0.02523876219638623, "p50_s": 0.0252655814983882, "p90_s": 0.02598772282944992, "p95_s": 0.026068960392149166, "p99_s": 0.02613395044230856, "min_s": 0.024300965014845133, "max_s": 0.02615019795484841 }, "linear.attn_qkv_proj.slice_reshape": { "count": 10, "mean_s": 5.790404975414276e-06, "p50_s": 5.527981556952e-06, "p90_s": 5.9999991208314885e-06, "p95_s": 7.080007344484326e-06, "p99_s": 7.9440139234066e-06, "min_s": 5.311972927302122e-06, "max_s": 8.160015568137169e-06 }, "attn_qkv_proj": { "count": 10, "mean_s": 0.0291815816948656, "p50_s": 0.029319992987439036, "p90_s": 0.02994130760198459, "p95_s": 0.029952906793914734, "p99_s": 0.02996218614745885, "min_s": 0.028194951999466866, "max_s": 0.02996450598584488 }, "attn_qkv_split_view": { "count": 10, "mean_s": 1.3921584468334913e-05, "p50_s": 1.2927979696542025e-05, "p90_s": 1.685120514594018e-05, "p95_s": 2.0033578039146953e-05, "p99_s": 2.257947635371238e-05, "min_s": 1.1119991540908813e-05, "max_s": 2.3215950932353735e-05 }, "attn_qk_rms_rope": { "count": 10, "mean_s": 0.012174421816598623, "p50_s": 0.01226081102504395, "p90_s": 0.01244935499271378, "p95_s": 0.012490539002465084, "p99_s": 0.012523486210266128, "min_s": 0.011791994038503617, "max_s": 0.01253172301221639 }, "attention_kernel": { "count": 10, "mean_s": 0.2819813063077163, "p50_s": 0.25794798001879826, "p90_s": 0.3529046454816125, "p95_s": 0.3818615222349762, "p99_s": 0.40502702363766735, "min_s": 0.25711314799264073, "max_s": 0.41081839898834005 }, "attn_output_reshape": { "count": 10, "mean_s": 4.5040040276944636e-06, "p50_s": 4.416011506691575e-06, "p90_s": 4.788796650245785e-06, "p95_s": 5.170408985577523e-06, "p99_s": 5.475698853842915e-06, "min_s": 4.064000677317381e-06, "max_s": 5.552021320909262e-06 }, "linear.attn_out_proj.flatten_contiguous": { "count": 10, "mean_s": 4.352000541985035e-06, "p50_s": 4.296016413718462e-06, "p90_s": 4.497583722695708e-06, "p95_s": 4.864789661951362e-06, "p99_s": 5.1585544133558866e-06, "min_s": 4.064000677317381e-06, "max_s": 5.231995601207018e-06 }, "linear.attn_out_proj.pre_quant_scale": { "count": 10, "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": 10, "mean_s": 9.830406634137035e-06, "p50_s": 9.512004908174276e-06, "p90_s": 1.074080355465412e-05, "p95_s": 1.1842395178973672e-05, "p99_s": 1.2723668478429319e-05, "min_s": 8.383998647332191e-06, "max_s": 1.2943986803293228e-05 }, "linear.attn_out_proj.bias_cast": { "count": 10, "mean_s": 3.606401151046157e-06, "p50_s": 3.6240089684724808e-06, "p90_s": 3.7360237911343575e-06, "p95_s": 3.772013587877154e-06, "p99_s": 3.8008054252713917e-06, "min_s": 3.4240074455738068e-06, "max_s": 3.8080033846199512e-06 }, "linear.attn_out_proj.activation_scale": { "count": 10, "mean_s": 0.0021395267103798687, "p50_s": 0.0021357620134949684, "p90_s": 0.0021502772287931293, "p95_s": 0.002160587624530308, "p99_s": 0.002168835941120051, "min_s": 0.002125058032106608, "max_s": 0.0021708980202674866 }, "linear.attn_out_proj.scale_to_device": { "count": 10, "mean_s": 1.1635292321443557e-05, "p50_s": 1.0199990356341004e-05, "p90_s": 1.3747194316238162e-05, "p95_s": 1.786559005267917e-05, "p99_s": 2.1160306641831997e-05, "min_s": 9.40798781812191e-06, "max_s": 2.1983985789120197e-05 }, "linear.attn_out_proj.activation_quant_pack": { "count": 10, "mean_s": 0.002853197598597035, "p50_s": 0.0028407784993760288, "p90_s": 0.0028848740214016287, "p95_s": 0.002903846013941802, "p99_s": 0.002919023607973941, "min_s": 0.0028314109658822417, "max_s": 0.0029228180064819753 }, "linear.attn_out_proj.activation_quant_wrap": { "count": 10, "mean_s": 6.711995229125023e-06, "p50_s": 6.647984264418483e-06, "p90_s": 7.030414417386055e-06, "p95_s": 7.275218376889824e-06, "p99_s": 7.471061544492841e-06, "min_s": 6.320013199001551e-06, "max_s": 7.520022336393595e-06 }, "linear.attn_out_proj.gemm": { "count": 10, "mean_s": 0.007781081803841516, "p50_s": 0.007702854520175606, "p90_s": 0.007810105266980827, "p95_s": 0.008148016140330581, "p99_s": 0.008418344839010387, "min_s": 0.007674487016629428, "max_s": 0.008485927013680339 }, "linear.attn_out_proj.slice_reshape": { "count": 10, "mean_s": 5.652802065014839e-06, "p50_s": 5.6480057537555695e-06, "p90_s": 5.96162280999124e-06, "p95_s": 6.004815804772079e-06, "p99_s": 6.03937020059675e-06, "min_s": 5.295965820550919e-06, "max_s": 6.0480087995529175e-06 }, "attn_out_proj": { "count": 10, "mean_s": 0.012856716598616914, "p50_s": 0.012750906986184418, "p90_s": 0.012935022206511349, "p95_s": 0.013290932588279247, "p99_s": 0.013575660893693565, "min_s": 0.012729579000733793, "max_s": 0.013646842970047146 }, "gate_msa": { "count": 10, "mean_s": 0.008448739902814851, "p50_s": 0.00843202299438417, "p90_s": 0.008597137464676052, "p95_s": 0.008597972249845043, "p99_s": 0.008598640077980235, "min_s": 0.008352135017048568, "max_s": 0.008598807035014033 }, "norm2": { "count": 10, "mean_s": 0.0034746349963825198, "p50_s": 0.0034624270047061145, "p90_s": 0.003538892592769116, "p95_s": 0.0035413837991654876, "p99_s": 0.003543376764282584, "min_s": 0.0034497790038585663, "max_s": 0.0035438750055618584 }, "modulate_mlp": { "count": 10, "mean_s": 0.010302344901720061, "p50_s": 0.0102713450032752, "p90_s": 0.010716759436763824, "p95_s": 0.010722224228084089, "p99_s": 0.010726596061140299, "min_s": 0.009972664003726095, "max_s": 0.010727689019404352 }, "linear.mlp_fc1.flatten_contiguous": { "count": 10, "mean_s": 4.3264124542474745e-06, "p50_s": 4.25602775067091e-06, "p90_s": 4.518410423770547e-06, "p95_s": 4.691208596341312e-06, "p99_s": 4.829447134397924e-06, "min_s": 4.096014890819788e-06, "max_s": 4.864006768912077e-06 }, "linear.mlp_fc1.pre_quant_scale": { "count": 10, "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": 10, "mean_s": 8.79999715834856e-06, "p50_s": 8.639995940029621e-06, "p90_s": 9.411235805600881e-06, "p95_s": 1.0073615703731773e-05, "p99_s": 1.0603519622236491e-05, "min_s": 8.03195871412754e-06, "max_s": 1.0735995601862669e-05 }, "linear.mlp_fc1.bias_cast": { "count": 10, "mean_s": 3.5087927244603633e-06, "p50_s": 3.5119883250445127e-06, "p90_s": 3.6495912354439497e-06, "p95_s": 3.6567944334819913e-06, "p99_s": 3.6625569919124246e-06, "min_s": 3.3599790185689926e-06, "max_s": 3.663997631520033e-06 }, "linear.mlp_fc1.activation_scale": { "count": 10, "mean_s": 0.0016255454160273076, "p50_s": 0.0016229460015892982, "p90_s": 0.0016354218998458237, "p95_s": 0.001644551966455765, "p99_s": 0.001651856019743718, "min_s": 0.0016106250113807619, "max_s": 0.0016536820330657065 }, "linear.mlp_fc1.scale_to_device": { "count": 10, "mean_s": 9.807990863919258e-06, "p50_s": 9.544019121676683e-06, "p90_s": 1.0929576819762588e-05, "p95_s": 1.1368762352503835e-05, "p99_s": 1.1720110778696835e-05, "min_s": 8.719973266124725e-06, "max_s": 1.1807947885245085e-05 }, "linear.mlp_fc1.activation_quant_pack": { "count": 10, "mean_s": 0.002176851505646482, "p50_s": 0.0021523534960579127, "p90_s": 0.0022936724068131297, "p95_s": 0.0022943492193007843, "p99_s": 0.0022948906692909074, "min_s": 0.002131377987097949, "max_s": 0.0022950260317884386 }, "linear.mlp_fc1.activation_quant_wrap": { "count": 10, "mean_s": 6.500800373032689e-06, "p50_s": 6.4159976318478584e-06, "p90_s": 6.894406396895647e-06, "p95_s": 7.175200153142213e-06, "p99_s": 7.399835158139467e-06, "min_s": 5.856039933860302e-06, "max_s": 7.455993909388781e-06 }, "linear.mlp_fc1.gemm": { "count": 10, "mean_s": 0.03409755709581077, "p50_s": 0.03422809299081564, "p90_s": 0.03452030192129314, "p95_s": 0.03457689344068058, "p99_s": 0.03462216665619053, "min_s": 0.03334671602351591, "max_s": 0.03463348496006802 }, "linear.mlp_fc1.slice_reshape": { "count": 10, "mean_s": 5.587207851931453e-06, "p50_s": 5.592009983956814e-06, "p90_s": 5.843234248459339e-06, "p95_s": 5.857640644535422e-06, "p99_s": 5.869165761396289e-06, "min_s": 5.296024028211832e-06, "max_s": 5.8720470406115055e-06 }, "mlp_fc1": { "count": 10, "mean_s": 0.037983797304332254, "p50_s": 0.038089904526714236, "p90_s": 0.03837030659778975, "p95_s": 0.03842020980373491, "p99_s": 0.03846013236849103, "min_s": 0.03735007898649201, "max_s": 0.03847011300968006 }, "mlp_swiglu": { "count": 10, "mean_s": 0.02382851129514165, "p50_s": 0.023903323977719992, "p90_s": 0.02399368488113396, "p95_s": 0.02401542845473159, "p99_s": 0.0240328233136097, "min_s": 0.023509315971750766, "max_s": 0.024037172028329223 }, "linear.mlp_fc2.flatten_contiguous": { "count": 10, "mean_s": 4.140805685892701e-06, "p50_s": 4.088011337444186e-06, "p90_s": 4.2416038922965525e-06, "p95_s": 4.392792470753193e-06, "p99_s": 4.513743333518505e-06, "min_s": 4.01597935706377e-06, "max_s": 4.543981049209833e-06 }, "linear.mlp_fc2.pre_quant_scale": { "count": 10, "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": 10, "mean_s": 8.01280839368701e-06, "p50_s": 7.863971404731274e-06, "p90_s": 8.312013233080506e-06, "p95_s": 8.996002725325523e-06, "p99_s": 9.54319431912154e-06, "min_s": 7.552036549896002e-06, "max_s": 9.679992217570543e-06 }, "linear.mlp_fc2.bias_cast": { "count": 10, "mean_s": 3.507203655317426e-06, "p50_s": 3.5119883250445127e-06, "p90_s": 3.576010931283235e-06, "p95_s": 3.6120007280260323e-06, "p99_s": 3.64079256542027e-06, "min_s": 3.4080003388226032e-06, "max_s": 3.6479905247688293e-06 }, "linear.mlp_fc2.activation_scale": { "count": 10, "mean_s": 0.004368299985071644, "p50_s": 0.004383971972856671, "p90_s": 0.004443581588566304, "p95_s": 0.004447440773947164, "p99_s": 0.004450528122251853, "min_s": 0.0042894280049949884, "max_s": 0.004451299959328026 }, "linear.mlp_fc2.scale_to_device": { "count": 10, "mean_s": 9.504001354798675e-06, "p50_s": 9.360024705529213e-06, "p90_s": 1.0358402505517006e-05, "p95_s": 1.0963209206238387e-05, "p99_s": 1.1447054566815497e-05, "min_s": 8.463975973427296e-06, "max_s": 1.1568015906959772e-05 }, "linear.mlp_fc2.activation_quant_pack": { "count": 10, "mean_s": 0.005924072209745645, "p50_s": 0.005892501038033515, "p90_s": 0.006118534551933408, "p95_s": 0.006128189770970493, "p99_s": 0.006135913946200162, "min_s": 0.005749636969994754, "max_s": 0.006137844990007579 }, "linear.mlp_fc2.activation_quant_wrap": { "count": 10, "mean_s": 6.358395330607891e-06, "p50_s": 6.351969204843044e-06, "p90_s": 6.553635466843843e-06, "p95_s": 6.74081384204328e-06, "p99_s": 6.89055654220283e-06, "min_s": 5.871988832950592e-06, "max_s": 6.927992217242718e-06 }, "linear.mlp_fc2.gemm": { "count": 10, "mean_s": 0.05625267179566436, "p50_s": 0.05627037599333562, "p90_s": 0.05668957601301372, "p95_s": 0.05669458000920713, "p99_s": 0.056698583206161855, "min_s": 0.055724094971083105, "max_s": 0.05669958400540054 }, "linear.mlp_fc2.slice_reshape": { "count": 10, "mean_s": 5.647999932989478e-06, "p50_s": 5.527981556952e-06, "p90_s": 5.806365516036748e-06, "p95_s": 6.375182420015334e-06, "p99_s": 6.830235943198204e-06, "min_s": 5.296024028211832e-06, "max_s": 6.943999323993921e-06 }, "mlp_fc2": { "count": 10, "mean_s": 0.06662256080890075, "p50_s": 0.06662932099425234, "p90_s": 0.06703189701656811, "p95_s": 0.06705320902692619, "p99_s": 0.06707025863521267, "min_s": 0.0659938960452564, "max_s": 0.06707452103728428 }, "gate_mlp": { "count": 10, "mean_s": 0.01003243929008022, "p50_s": 0.008418151992373168, "p90_s": 0.010617988393642003, "p95_s": 0.017215404182206823, "p99_s": 0.022493336813058706, "min_s": 0.008402118983212858, "max_s": 0.02381281997077167 }, "block_total": { "count": 10, "mean_s": 0.510889746597968, "p50_s": 0.48557983900536783, "p90_s": 0.5824499189096968, "p95_s": 0.6108186564553761, "p99_s": 0.6335136464919197, "min_s": 0.483186541008763, "max_s": 0.6391873940010555 } }, "module_forward": { "count": 10, "mean_s": 0.4524989926023409, "p50_s": 0.4523211540072225, "p90_s": 0.4534270477830432, "p95_s": 0.4535232613969129, "p99_s": 0.4536002322880086, "min_s": 0.4518661299953237, "max_s": 0.4536194750107825 }, "module_forward_checksum": -23765568.0, "fused_elementwise": true, "profiler_summary": { "profiled_iterations": 1, "runtime_kernel_launches": 0, "runtime_kernel_launches_per_block": 0.0, "positive_self_device_allocated_bytes": 8009578496 }, "profiler_top_events": [ { "key": "aten::rms_norm", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 1857.8570000000002, "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": 1847.825, "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": 966.9929999999995, "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": 728.1770000000006, "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": "aten::contiguous", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 447.7280000000005, "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": 438.51199999999926, "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": 437.3929999999996, "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": 44, "device_type": "DeviceType.CPU", "cpu_time_total_us": 406.1919999999975, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 8009561088, "input_shapes": "[[], [], [], [], [], []]" }, { "key": "aten::zero_", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 388.7049999999999, "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::rms_rope_split_half_", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 314.2399999999998, "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::copy_", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 111.39199999999983, "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": "aten::linear", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 109.80799999999999, "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::linear", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 96.46399999999994, "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": 92.30400000000009, "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::linear", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 88.17599999999948, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 14336], [5376, 14336], []]" }, { "key": "aten::mean", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 83.29599999999937, "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": "comfy_kitchen::scaled_mm_nvfp4", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 81.71200000000044, "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": "comfy_kitchen::quantize_nvfp4", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 80.55999999999949, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810, 7168], [], [], [], []]" }, { "key": "comfy_kitchen::scaled_mm_nvfp4", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 76.44799999999941, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": -512, "input_shapes": "[[37824, 7168], [5376, 7168], [], [], [37888, 896], [5376, 896], [], [], []]" }, { "key": "aten::mul", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 58.944000000001324, "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": 37.712000000000444, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[], [], [], [], [], [], []]" }, { "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": 33.90399999999954, "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::split", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 32.55999999999949, "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::fill_", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 30.048999999999978, "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::to", "count": 9, "device_type": "DeviceType.CPU", "cpu_time_total_us": 27.99999999999909, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[], [], [], [], []]" }, { "key": "aten::empty_like", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 26.623999999999796, "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::copy_", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 23.200000000000728, "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": 20.592000000000553, "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": 16.384000000000015, "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::to", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 16.1279999999997, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[], [], [], [], [], []]" }, { "key": "aten::squeeze", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 9.936000000000604, "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::reshape", "count": 5, "device_type": "DeviceType.CPU", "cpu_time_total_us": 9.615999999999985, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[], []]" }, { "key": "aten::zero_", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 8.351999999999862, "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::view", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 8.288000000000011, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[37810], []]" }, { "key": "aten::empty_strided", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 6.608000000001084, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 1024, "input_shapes": "[[], [], [], [], [], []]" }, { "key": "aten::zero_", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 6.527999999999338, "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::fill_", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 6.431999999999789, "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": 6.11200000000008, "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::view", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 5.711999999999534, "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::as_strided", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 5.520000000000437, "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::to", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 5.519999999999413, "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": 5.1840000000001965, "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::slice", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 3.6479999999992287, "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::as_strided", "count": 3, "device_type": "DeviceType.CPU", "cpu_time_total_us": 2.6879999999991924, "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::view", "count": 5, "device_type": "DeviceType.CPU", "cpu_time_total_us": 2.496000000001004, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[], []]" }, { "key": "aten::view", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 2.4639999999990323, "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": 2.319999999999709, "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": 1.9520000000002256, "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::as_strided", "count": 4, "device_type": "DeviceType.CPU", "cpu_time_total_us": 1.5679999999993015, "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": 1.4880000000002838, "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": 1.2960000000002765, "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::reshape", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 1.1520000000000437, "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": 1.0399999999999636, "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": 2, "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": "[[37824, 5376], [], [], []]" }, { "key": "aten::view", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.8000000000001819, "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::alias", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.7839999999996508, "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::alias", "count": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.7200000000002547, "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": 2, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.5280000000002474, "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::view", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.4639999999999418, "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": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.44800000000032014, "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::as_strided", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.431999999999789, "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::as_strided", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.431999999999789, "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::alias", "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, 28672]]" }, { "key": "aten::view", "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, 5376], []]" }, { "key": "aten::view", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.28800000000046566, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[5376, 896], []]" }, { "key": "aten::view", "count": 1, "device_type": "DeviceType.CPU", "cpu_time_total_us": 0.28799999999955617, "device_time_total_us": 0.0, "self_device_time_total_us": 0.0, "self_device_memory_usage_bytes": 0, "input_shapes": "[[5376, 448], []]" }, { "key": "[memory]", "count": 36, "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": -8009273344, "input_shapes": "[]" } ] }