CPU Tasks LoadModel 3.22847 seconds RunComplete 14.2729 seconds Run 14.186 seconds Callbacks 674.7 microseconds, 4 calls, 168.675 microseconds average Spectrogram 29.6112 milliseconds, 3 calls, 9.8704 milliseconds average Sample 7.7473 milliseconds, 27 calls, 286.937 microseconds average Encode 11.8931 seconds Decode 2.29185 seconds DecodeStep 2.28406 seconds, 27 calls, 84.5949 milliseconds average GPU Tasks LoadModel 1.9083 seconds Run 14.0698 seconds Encode 11.9404 seconds EncodeLayer 10.2786 seconds, 32 calls, 321.205 milliseconds average Decode 2.12941 seconds DecodeStep 2.12938 seconds, 27 calls, 78.8661 milliseconds average DecodeLayer 1.98655 seconds, 864 calls, 2.29924 milliseconds average Compute Shaders mulMatTiledEx 8.49883 seconds, 320 calls, 26.5589 milliseconds average mulMatTiled 2.04655 seconds, 385 calls, 5.31573 milliseconds average mulMatByRowTiled 982.48 milliseconds, 8346 calls, 117.719 microseconds average mulMatByRowTiledEx 481.123 milliseconds, 1664 calls, 289.137 microseconds average convolutionMain2Fixed 415.244 milliseconds softMaxFixed 404.223 milliseconds, 896 calls, 451.142 microseconds average matReshapePanels 210.915 milliseconds, 193 calls, 1.09282 milliseconds average norm 154.589 milliseconds, 2684 calls, 57.5965 microseconds average convolutionMain 126.883 milliseconds addRepeatGelu 112.131 milliseconds, 898 calls, 124.867 microseconds average copyConvert 100.589 milliseconds, 1856 calls, 54.1968 microseconds average scaleInPlace 91.3539 milliseconds, 896 calls, 101.957 microseconds average fmaRepeat1 86.7731 milliseconds, 2684 calls, 32.3298 microseconds average copyTranspose 77.5852 milliseconds, 1792 calls, 43.2953 microseconds average addRepeat 76.3677 milliseconds, 960 calls, 79.5497 microseconds average addRepeatEx 70.8699 milliseconds, 2656 calls, 26.6829 microseconds average softMaxLong 39.4035 milliseconds, 27 calls, 1.45939 milliseconds average addRepeatScale 25.0842 milliseconds, 1728 calls, 14.5163 microseconds average softMax 14.0555 milliseconds, 864 calls, 16.2679 microseconds average convolutionPrep1 13.9331 milliseconds, 2 calls, 6.96655 milliseconds average diagMaskInf 8.1717 milliseconds, 864 calls, 9.45799 microseconds average convolutionPrep2 5.2098 milliseconds, 2 calls, 2.6049 milliseconds average add 2.9724 milliseconds addRows 91.2 microseconds, 27 calls, 3.37778 microseconds average Memory Usage Model 892.591 KB RAM, 2.8815 GB VRAM Context 1.98376 MB RAM, 1.13447 GB VRAM Total 2.85543 MB RAM, 4.01597 GB VRAM