我是 Qualcomm AI Engine Direct SDK (QNN) 的新手。在直接将 AI 模型部署到高通设备之前,我想看一下 QNN 的 x86_64 后端,这也与 QNN 的量化过程相关。
但是,我发现Qualcomm的QNN x86_64 CPU后端88x比inception_v3的onnxruntime慢。
以下是重现该问题的步骤:
按照Qualcomm的指示安装QNN SDK。
下载模型并转换为ONNX
import torch
# Source of model: https://pytorch.org/hub/pytorch_vision_inception_v3/
model = torch.hub.load("pytorch/vision:v0.10.0", "inception_v3", pretrained=True)
model.eval()
x = torch.rand(1, 3, 299, 299)
torch.onnx.export(model, x, "inception_v3.onnx", opset_version=17)
${QNN_SDK_ROOT}/bin/x86_64-linux-clang/qnn-onnx-converter \
--input_network inception_v3.onnx \
--input_dim 'x.1' 1,3,299,299 \
--out_node '914' \
--output_path inception_v3.cpp
mkdir -p model_libs
${QNN_SDK_ROOT}/bin/x86_64-linux-clang/qnn-model-lib-generator \
-c inception_v3.cpp \
-b inception_v3.bin \
-t x86_64-linux-clang \
-o model_libs
import numpy as np
np.random.rand(3, 299, 299).astype(np.float32).tofile("input.raw")
# see rest in Ref: https://github.com/quic/ai-hub-models/issues/17
和
echo input.raw > input.txt
${QNN_SDK_ROOT}/bin/x86_64-linux-clang/qnn-net-run \
--backend ${QNN_SDK_ROOT}/lib/x86_64-linux-clang/libQnnCpu.so \
--model model_libs/x86_64-linux-clang/libinception_v3.so \
--input_list input.txt \
--profiling_level=basic \
--keep_num_outputs=0 \
--num_inferences=10
并查看日志
${QNN_SDK_ROOT}/bin/x86_64-linux-clang/qnn-profile-viewer --input_log output/qnn-profiling-data_0.log
这是日志
Input Log File Location: output/qnn-profiling-data_0.log
Log File Created: Thu Sep 26 08:49:41 2024
Time Scale: 1e-06
Epoch Timestamp: 1727340581547093 Steady Clock Timestamp: 1380276319731
Generated using:
qnn-profile-viewer v2.26.0.240827110523_99241
qnn-net-run v2.26.0.240827110523_99241
Backend v2.26.0.240827110523_99241
Qnn Init/Prepare/Finalize/De-Init/Execute/Lib-Load Statistics:
------------------------------------------------------------
Init Stats:
-----------
NetRun: 171679 us
Compose Graphs Stats:
--------------
NetRun: 95902 us
Finalize Stats:
---------------
Graph 0 (inception_v3):
NetRun: 75775 us
Backend (GRAPH_FINALIZE): 75769 us
De-Init Stats:
--------------
NetRun: 20778 us
Backend (null): 0 us
Execute Stats (Overall):
------------------------
NetRun IPS (includes IO and misc. time): 0.5542 inf/sec
Execute Stats (Average):
------------------------
Total Inference Time:
---------------------
Graph 0 (inception_v3):
NetRun: 1803480 us
Backend (GRAPH_EXECUTE): 1803294 us
Execute Stats (Min):
------------------------
Total Inference Time:
---------------------
Graph 0 (inception_v3):
NetRun: 1754020 us
Backend (GRAPH_EXECUTE): 1753902 us
Execute Stats (Max):
------------------------
Total Inference Time:
---------------------
Graph 0 (inception_v3):
NetRun: 1895948 us
Backend (GRAPH_EXECUTE): 1895815 us
我们看到了
Backend (GRAPH_EXECUTE): 1803294 us
。但是,由 ONNXRuntime CPU 提供商运行相同的 onnx
import numpy as np
import onnxruntime
x = np.random.rand(1, 3, 299, 299).astype(np.float32)
session = onnxruntime.InferenceSession(
"inception_v3.onnx", providers=["CPUExecutionProvider"]
)
outputs = session.run(["914"], input_feed={"x.1": x})
import time
N = 100
t1 = time.time()
for _ in range(N):
outputs = session.run(["914"], input_feed={"x.1": x})
t2 = time.time()
print(f"average inference time = {(t2 - t1)/N*1000} miliseconds")
输出是
average inference time = 21.910243034362793 miliseconds
所以我想知道为什么 QNN 的 x86_64 CPU 后端明显比 onnxruntime 慢(1803.294 ms vs 21.91ms)?
如有任何帮助,我们将不胜感激。
PS
62.1ms [ INFO ] [QNN_CPU] QnnGraph execute start
2086.4ms [ INFO ] [QNN_CPU] QnnGraph execute end
这种缓慢的执行时间与上面 QNN 的执行时间一致。
-O3
被添加到临时文件夹中的CXX_FLAGS
中的Makefile.linux-x86_64
中,由qnn-model-lib-generator
生成。我尝试了和你一样的方法,得到的一切几乎与onnx模型相同,即28ms