- Sort Score
- Result 10 results
- Languages All
Results 1 - 1 of 1 for 712 (0.04 sec)
-
tensorflow/compiler/mlir/lite/stablehlo/transforms/compose_uniform_quantized_type_pass.cc
// %10 = stablehlo.convert %9 // i8 -> f32 cast trick for z1. // %11 = stablehlo.convert %5 // i8 -> f32 cast trick for filter. // %12 = stablehlo.convolution(%10, %11) // q2 * z1 // %13 = stablehlo.subtract %7, %12 // q1 * q2 - q2 * z1 // %14 = stablehlo.constant // Merged scale s1 * s2, precalculated. // %15 = stablehlo.broadcast_in_dim %14 // %16 = stablehlo.multiply %13 %15 // r3 = s1 s2 (q1 q2 - q2 z1) //
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu Apr 25 16:01:03 UTC 2024 - 64.6K bytes - Viewed (0)