alderlake.txt

cmake -GNinja -B ../iree-build/ -S . -DCMAKE_BUILD_TYPE=RelWithDebInfo -DIREE_ENABLE_ASSERTIONS=OFF -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ -DCMAKE_INSTALL_PREFIX=/home/chi/IREE/iree-build/install -DIREE_ENABLE_LLD=ON -DIREE_HAL_DRIVER_VULKAN=ON -DIREE_TARGET_BACKEND_CUDA=OFF -DIREE_TARGET_BACKEND_VULKAN_SPIRV=ON -DIREE_TARGET_BACKEND_OPENCL_SPIRV=ON -DIREE_ENABLE_ASSERTIONS=ON -DIREE_BUILD_PYTHON_BINDINGS=ON -DPython3_EXECUTABLE="$(which python)" -DIREE_ENABLE_RUNTIME_TRACING=ON -DIREE_BYTECODE_MODULE_FORCE_LLVM_SYSTEM_LINKER=ON -DIREE_BUILD_TRACY=ON 

iree-compile --iree-input-type=none --iree-hal-target-backends=vulkan --iree-vulkan-target-triple=rdna2-6900xt-linux --iree-stream-resource-index-bits=64 --iree-vm-target-index-bits=64  unet_stable_diff_maxf.mlir -o unet_stable_diff_maxf.vmfb

iree-benchmark-module --module_file=unet_stable_diff_maxf.vmfb --entry_function=forward --device=vulkan --function_input="2x64x64x4xf32" --function_input="2x320xf32" --function_input="2x77x768xf32"

for tf model desktop

iree-compile --iree-input-type=mhlo --iree-hal-target-backends=llvm-cpu --iree-stream-resource-index-bits=64 --iree-vm-target-index-bits=64 stable_diff_tf.mlir -o stable_diff_tf.vmfb

iree-benchmark-module --module_file=stable_diff_tf.vmfb --entry_function=forward --device=local-task --function_input="2x64x64x4xf32" --function_input="2x320xf32" --function_input="2x77x768xf32"

Terminal 1 on desktop
TRACY_NO_EXIT=1 IREE_PRESERVE_DYLIB_TEMP_FILES=1 iree-benchmark-module --module_file=stable_diff_tf.vmfb --entry_function=forward --device=local-task --function_input="2x64x64x4xf32" --function_input="2x320xf32" --function_input="2x77x768xf32"
Terminal 2 on desktop
tracy/iree-tracy-capture -o myprofile.tracy

on local ubuntu desktop
scp chi@arc:/home/chi/IREE/myprofile.tracy /home/chi/src/ubuntu20/shark/IREE/vulkan
tracy/iree-tracy-profiler myprofile.tracy

scp /home/chi/src/ubuntu20/shark/IREE/shark-tuner.tar.gz chi@arc:/home/chi/IREE/

python shark-tuner/minilm_example.py -model /home/chi/IREE/stable_diff_tf.mlir -num_iters 10 -result_dir results -device gpu -search_op matmul

python shark-tuner/minilm_example.py -model /home/chi/IREE/unet_stable_diff_maxf.mlir -num_iters 600 -result_dir results -device cpu -search_op conv

unet_stable_diff_maxf.mlir Bug Print

(tuner_venv) chi@alderlake:~/IREE$ python shark-tuner/minilm_example.py -model /home/chi/IREE/unet_stable_diff_maxf.mlir -num_iters 600 -result_dir results -device cpu -search_op conv
The input mlir type is 
Traceback (most recent call last):
  File "/home/chi/IREE/shark-tuner/minilm_example.py", line 32, in <module>
    tuner.export_tuned_model()
  File "/home/chi/IREE/shark-tuner/shark_tuner/shark_tuner.py", line 78, in export_tuned_model
    model_config_path = self.save_model_configs()
  File "/home/chi/IREE/shark-tuner/shark_tuner/shark_tuner.py", line 72, in save_model_configs
    model_config_path = generate_model_configs(self.search_config_path,
  File "/home/chi/IREE/shark-tuner/shark_tuner/tuner_utils.py", line 237, in generate_model_configs
    with open(log_path, 'r') as f:
FileNotFoundError: [Errno 2] No such file or directory: '/home/chi/IREE/results/best_configs.json'
(tuner_venv) chi@alderlake:~/IREE$ python shark-tuner/minilm_example.py -model /home/chi/IREE/unet_stable_diff_maxf.mlir -num_iters 600 -result_dir results -device cpu -search_op matmul
The input mlir type is linalg
Searching for [0, 2, 1280, 320]
Updated op %2 = linalg.matmul {compilation_info = #iree_codegen.compilation_info<lowering_config = <tile_sizes = [[2, 6], [1, 1, 0], [0, 0, 32]]>, translation_info = <CPUDoubleTilingPadExpert>>} ins(%arg0, %arg1 : tensor<2x320xf32>, tensor<320x1280xf32>) outs(%1 : tensor<2x1280xf32>) -> tensor<2x1280xf32>
^Ciree-benchmark-module Error ['iree-benchmark-module', '--module_file=/home/chi/tmp/matmul_2x1280x320.vmfb', '--entry_function=matmul', '--device=local-task', '--function_input=2x320xf32', '--function_input=320x1280xf32', '--batch_size=100', "| awk 'END{{print $2 $3}}'"]
Updated op %2 = linalg.matmul {compilation_info = #iree_codegen.compilation_info<lowering_config = <tile_sizes = [[2, 4], [1, 3, 0], [0, 0, 12]]>, translation_info = <CPUDoubleTilingPadExpert>>} ins(%arg0, %arg1 : tensor<2x320xf32>, tensor<320x1280xf32>) outs(%1 : tensor<2x1280xf32>) -> tensor<2x1280xf32>

Sharktunner

(tuner_venv) chi@alderlake:~/IREE$ python shark-tuner/minilm_example.py -model /home/chi/IREE/stable_diff_linalg.mlir -num_iters 100 -result_dir results -device vulkan -search_op conv
The input mlir type is linalg
Found AMD Radeon RX 5000 series device. Using rdna1-5700xt-linux
Searching for [2, 66, 66, 4, 3, 3, 320, 64, 64, 1, 1, 0]
Updated op %2 = linalg.conv_2d_nhwc_hwcf {compilation_info = #iree_codegen.compilation_info<lowering_config = <tile_sizes = [[0, 64, 64, 16], [0, 4, 8, 4], [0, 0, 0, 0, 1, 1, 4], [0, 1, 0, 0]]>, translation_info = <SPIRVVectorize>, workgroup_size = [4, 8, 16]>, dilations = dense<1> : tensor<2xi64>, strides = dense<1> : tensor<2xi64>} ins(%arg0, %arg1 : tensor<2x66x66x4xf32>, tensor<3x3x4x320xf32>) outs(%1 : tensor<2x64x64x320xf32>) -> tensor<2x64x64x320xf32>