{ "environment": { "platform": "macOS-27.0.1-arm64-arm-64bit", "machine": "arm64", "metal": { "device_name": "Apple M5 Pro", "max_recommended_working_set_size": 55662788608, "memory_size": 68719476736, "architecture": "applegpu_g17s", "max_buffer_length": 41747087360, "resource_limit": 499000 }, "versions": { "mlx": "0.32.4", "mlx-metal": "0.32.4", "numpy": "2.5.4", "safetensors": "0.8.0", "transformers": "4.57.6", "torch": "2.15.0", "Pillow": "13.0.0", "scikit-image": "0.26.0" }, "precision_environment": { "MLX_ENABLE_TF32": "0", "MLX_CONV_WINOGRAD": "0" } }, "trace": "outputs/suite/product/native-3.trace.safetensors", "trials": [ { "group": "qk_native_reduce", "head_major": false, "max_abs": 0.0, "relative_rmse": 0.0, "equal_fraction": 1.0, "order": [ "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control" ], "control": { "median_ms": 9.447520948015153, "min_ms": 9.296874981373549, "max_ms": 9.703416842967272, "samples_ms": [ 9.703416842967272, 9.584542131051421, 9.37366602011025, 9.409416932612658, 9.487750008702278, 9.296874981373549, 9.340708842501044, 9.48562496341765, 9.311167057603598, 9.512332966551185 ] }, "candidate": { "median_ms": 1.164812478236854, "min_ms": 1.1362920049577951, "max_ms": 1.3076670002192259, "samples_ms": [ 1.2030829675495625, 1.1704580392688513, 1.1736671440303326, 1.1448750738054514, 1.137667102739215, 1.145332818850875, 1.1591669172048569, 1.1362920049577951, 1.2299581430852413, 1.3076670002192259 ] }, "reduction_pct": 87.67070764228819 }, { "group": "qk_native_reduce", "head_major": true, "max_abs": 0.0, "relative_rmse": 0.0, "equal_fraction": 1.0, "order": [ "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control" ], "control": { "median_ms": 9.508854593150318, "min_ms": 9.290958987548947, "max_ms": 9.688917081803083, "samples_ms": [ 9.504167130216956, 9.363082936033607, 9.290958987548947, 9.551124880090356, 9.65708284638822, 9.688917081803083, 9.496458107605577, 9.534791111946106, 9.51354205608368, 9.456166066229343 ] }, "candidate": { "median_ms": 1.2558334274217486, "min_ms": 1.1507498566061258, "max_ms": 1.3518338091671467, "samples_ms": [ 1.1622910387814045, 1.1507498566061258, 1.1727078817784786, 1.1655841954052448, 1.3518338091671467, 1.268249936401844, 1.324875047430396, 1.3020830228924751, 1.3239579275250435, 1.2434169184416533 ] }, "reduction_pct": 86.79301050279615 }, { "group": "fc2_compiled_block", "max_abs": 0.0, "relative_rmse": 0.0, "equal_fraction": 1.0, "order": [ "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control" ], "control": { "median_ms": 183.73187503311783, "min_ms": 179.05995808541775, "max_ms": 187.52654106356204, "samples_ms": [ 179.05995808541775, 180.07337511517107, 180.4345001000911, 180.9190409258008, 182.82225006259978, 184.64150000363588, 185.34387508407235, 186.96787487715483, 186.44287483766675, 187.52654106356204 ] }, "candidate": { "median_ms": 178.05443797260523, "min_ms": 172.12808318436146, "max_ms": 182.86733282729983, "samples_ms": [ 172.12808318436146, 174.88345806486905, 175.65049999393523, 176.61041696555912, 178.7279169075191, 177.38095903769135, 178.76341682858765, 182.4859578628093, 182.64337512664497, 182.86733282729983 ] }, "reduction_pct": 3.090066467502839 }, { "group": "qk_compiled_block", "max_abs": 0.0, "relative_rmse": 0.0, "equal_fraction": 1.0, "order": [ "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control" ], "control": { "median_ms": 194.46229189634323, "min_ms": 189.9893330410123, "max_ms": 199.30545799434185, "samples_ms": [ 189.9893330410123, 191.2787079345435, 192.03466712497175, 193.5009998269379, 192.95945903286338, 195.42358396574855, 196.06483285315335, 197.53999984823167, 197.66208296641707, 199.30545799434185 ] }, "candidate": { "median_ms": 189.1520629869774, "min_ms": 184.64308395050466, "max_ms": 192.54904193803668, "samples_ms": [ 184.72437490709126, 184.64308395050466, 186.77700008265674, 186.95037486031651, 189.0332920011133, 189.2708339728415, 191.06099987402558, 190.72229112498462, 192.54904193803668, 191.98762509040534 ] }, "reduction_pct": 2.730724222975023 }, { "group": "qk_and_fc2_compiled_block", "max_abs": 0.0, "relative_rmse": 0.0, "equal_fraction": 1.0, "order": [ "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control", "control", "candidate", "candidate", "control" ], "control": { "median_ms": 205.9104380896315, "min_ms": 201.97012485004961, "max_ms": 211.21824998408556, "samples_ms": [ 201.97012485004961, 202.7630410157144, 203.4202078357339, 205.13020898215473, 205.30466712079942, 206.9944590330124, 206.51620905846357, 208.5858341306448, 209.9133748561144, 211.21824998408556 ] }, "candidate": { "median_ms": 195.80483296886086, "min_ms": 192.06016696989536, "max_ms": 199.2452498525381, "samples_ms": [ 192.79541703872383, 192.06016696989536, 193.49937490187585, 194.14429203607142, 195.60312503017485, 196.824082871899, 196.68316701427102, 196.00654090754688, 199.08954203128815, 199.2452498525381 ] }, "reduction_pct": 4.9077672868491184 } ], "memory": { "sampled_peak_physical_bytes": 3273049384, "lifetime_peak_physical_bytes": 3546989816, "sample_interval_s": 0.1, "sample_count": 146, "scope": "Parent process only; includes Metal; excludes compiler XPC services. Counters overlap; do not sum." } }