CPU Tasks
LoadModel       7.92578 seconds
RunComplete     8.33686 seconds
Run     8.25683 seconds
Callbacks       337.7 microseconds, 4 calls, 84.425 microseconds average
Spectrogram     16.4214 milliseconds, 3 calls, 5.4738 milliseconds average
Sample  3.8768 milliseconds, 27 calls, 143.585 microseconds average
Encode  4.14309 seconds
Decode  4.11338 seconds
DecodeStep      4.10947 seconds, 27 calls, 152.203 milliseconds average
    GPU Tasks
LoadModel       7.53025 seconds
Run     8.05464 seconds
Encode  3.98133 seconds
EncodeLayer     3.29696 seconds, 32 calls, 103.03 milliseconds average
Decode  4.07331 seconds
DecodeStep      4.07331 seconds, 27 calls, 150.863 milliseconds average
DecodeLayer     3.81856 seconds, 864 calls, 4.41963 milliseconds average
    Compute Shaders
mulMatTiled     3.6307 seconds, 705 calls, 5.14993 milliseconds average
mulMatByRowTiled        1.45034 seconds, 10010 calls, 144.889 microseconds average
fmaRepeat1      774.011 milliseconds, 2684 calls, 288.38 microseconds average
copyTranspose   416.996 milliseconds, 1792 calls, 232.699 microseconds average
addRepeatScale  223.798 milliseconds, 1728 calls, 129.513 microseconds average
norm    211.821 milliseconds, 2684 calls, 78.9197 microseconds average
softMaxFixed    209.124 milliseconds, 896 calls, 233.398 microseconds average
copyConvert     186.898 milliseconds, 1856 calls, 100.699 microseconds average
addRepeatEx     176.985 milliseconds, 2656 calls, 66.6361 microseconds average
softMaxLong     162.482 milliseconds, 27 calls, 6.01787 milliseconds average
convolutionMain2Fixed   154.739 milliseconds
addRepeatGelu   132.088 milliseconds, 898 calls, 147.091 microseconds average
softMax 98.1905 milliseconds, 864 calls, 113.646 microseconds average
scaleInPlace    57.3956 milliseconds, 896 calls, 64.0576 microseconds average
convolutionMain 45.7933 milliseconds
addRepeat       36.9315 milliseconds, 960 calls, 38.4703 microseconds average
diagMaskInf     28.2987 milliseconds, 864 calls, 32.7531 microseconds average
convolutionPrep1        9.0334 milliseconds, 2 calls, 4.5167 milliseconds average
convolutionPrep2        1.1608 milliseconds, 2 calls, 580.4 microseconds average
add     320 microseconds
addRows 244 microseconds, 27 calls, 9.03704 microseconds average
    Memory Usage
Model   892.591 KB RAM, 2.8815 GB VRAM
Context 1.98376 MB RAM, 1.07641 GB VRAM
Total   2.85543 MB RAM, 3.95791 GB VRAM