mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-05-11 17:00:18 +00:00
4ef704d8a6
* Add support for mixed precision in contraction scale and bilinear (#936) * Extract common functionality to separate files * Reference contraction: Remove incorrect consts from type_converts * Reference contraction: Add missing type_convert for dst value * Reference contraction: Fix incorrect order of B matrix dimensions * Add support for mixed precision in contraction scale and bilinear * Move using statements from instances to a common file * Move using statements from examples to a common file * Fix the order of B matrix dimensions across examples and profiler * Fix the computation of error threshold * Make ComputeDataType an optional argument * Include possible DataType -> ComputeDataType casting error in the threshold * Remove commented code * Make the ComputeDataType an optional argument in instance --------- Co-authored-by: Illia Silin <98187287+illsilin@users.noreply.github.com>
Instructions for example_contraction_bilinear_xdl_fp32
Run
#arg1: verification (0=no, 1=yes)
#arg2: initialization (0=no init, 1=integer value, 2=decimal value)
#arg3: time kernel (0=no, 1=yes)
./bin/example_contraction_bilinear_xdl_fp32 1 1 1
Result (MI100 @ dynammic freq, 46TFlops peak FP32)
a_ms_ks: dim 4, lengths {30, 128, 32, 64}, strides {524288, 4096, 128, 1}
b_ks_ns: dim 4, lengths {32, 64, 32, 64}, strides {128, 1, 524288, 4096}
c_ms_ns: dim 4, lengths {30, 128, 32, 64}, strides {524288, 4096, 128, 1}
launch_and_time_kernel: grid_dim {240, 1, 1}, block_dim {256, 1, 1}
Warm up 1 time
Start running 10 times...
Perf: 0.843286 ms, 38.1985 TFlops, 94.5014 GB/s, DeviceContractionMultipleD_Xdl_CShuffle<256, 256, 128, 16, 4, 4>