网络建设网站,wordpress修改主题目录,ajax网站开发技术,wordpress增加用户活跃RK3588-NPU pytorch-image-models 模型编译测试 一.背景二.操作步骤1.下载依赖2.创建容器3.安装依赖4.创建脚本A.生成模型名列表B.生成ONNX模型C.生成RKNN模型D.批量测试脚本 一.背景
测试RK3588-NPU对https://github.com/huggingface/pytorch-image-models.git中模型的支持程… RK3588-NPU pytorch-image-models 模型编译测试 一.背景二.操作步骤1.下载依赖2.创建容器3.安装依赖4.创建脚本A.生成模型名列表B.生成ONNX模型C.生成RKNN模型D.批量测试脚本 一.背景
测试RK3588-NPU对https://github.com/huggingface/pytorch-image-models.git中模型的支持程度
二.操作步骤
1.下载依赖
mkdir rk3588
cd rk3588
wget -O v2.3.0.tar.gz https://github.com/airockchip/rknn-toolkit2/archive/refs/tags/v2.3.0.tar.gz
tar -xf v2.3.0.tar.gz
git clone https://github.com/huggingface/pytorch-image-models.git2.创建容器
docker stop rk3588_npu_test
docker rm rk3588_npu_test
docker run --gpus all --shm-size128g -id -e NVIDIA_VISIBLE_DEVICESall \--privileged --nethost -v $PWD:/home -w /home \--namerk3588_npu_test nvcr.io/nvidia/pytorch:23.07-py3 /bin/bash
docker start rk3588_npu_test
docker exec -ti rk3588_npu_test bash3.安装依赖
cd /home
pip install -r rknn-toolkit2-2.3.0/rknn-toolkit2/packages/x86_64/requirements_cp310-2.3.0.txt -i https://pypi.tuna.tsinghua.edu.cn/simple/
pip install rknn-toolkit2-2.3.0/rknn-toolkit2/packages/x86_64/rknn_toolkit2-2.3.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
cd /home/pytorch-image-models
pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple/
pip uninstall opencv-contrib-python opencv-python -y
pip install opencv-python4.7.0.72 -i https://pypi.tuna.tsinghua.edu.cn/simple/
pip install opencv-contrib-python4.7.0.72 -i https://pypi.tuna.tsinghua.edu.cn/simple/4.创建脚本
A.生成模型名列表
cat list_models.py -EOF
from timm.models import create_model, is_model, list_models
model_names list_models()
with open(models.txt,w) as f:for n in model_names:f.write(f{n}\n)
EOF
python list_models.pyB.生成ONNX模型
cat gen_onnx.py-EOF
import timm
import sys
import torch
import os
from timm.utils.model import reparameterize_model
from timm.utils.onnx import onnx_exportONNX_MODEL model.onnx
if os.path.exists(ONNX_MODEL):os.remove(ONNX_MODEL)
model timm.create_model(sys.argv[1],num_classes2,in_chans3,pretrainedFalse,exportableTrue,
)
model.eval()
input_sizemodel.default_cfg.get(input_size)
input_tensor torch.ones((1,) input_size)
input_names [input]
output_names [output]
torch.onnx.export(model, input_tensor, ONNX_MODEL, verboseFalse, input_namesinput_names,output_namesoutput_names,opset_version17,export_paramsTrue)
EOFC.生成RKNN模型
cat gen_rknn.py-EOF
import os
import urllib
import traceback
import time
import sys
import numpy as np
import cv2
from rknn.api import RKNNONNX_MODEL model.onnx
RKNN_MODEL model.rknnif not os.path.exists(ONNX_MODEL):exit(-1)if os.path.exists(RKNN_MODEL):os.remove(RKNN_MODEL)
if not os.path.exists(./dataset.txt):imgnp.ones((224,224,3),dtypenp.int8)cv2.imwrite(img.jpg,img)with open(./dataset.txt,w) as f:f.write(img.jpg)rknn RKNN(verboseFalse)
rknn.config(mean_values[123.675, 116.28, 103.53], std_values[58.82, 58.82, 58.82], target_platformrk3588)
ret rknn.load_onnx(modelONNX_MODEL)
if ret ! 0:exit(ret)
ret rknn.build(do_quantizationTrue, dataset./dataset.txt)
if ret ! 0:exit(ret)
ret rknn.export_rknn(RKNN_MODEL)
if ret ! 0:exit(ret)
rknn.release()
if os.path.exists(RKNN_MODEL):print(BUILD SUCCESS)
EOFD.批量测试脚本
cat run.sh-EOF
#!/bin/bashmkdir -p ./log
# 获取 models.txt 中的模型总数
total$(wc -l models.txt)
count0
for name in $(cat models.txt); doif [ ! -f ./log/$name ]; then# 打印进度条progress$((count * 100 / total))echo -ne 进度: [for ((i 0; i progress; i2)); do echo -n ; donefor ((i progress; i 100; i2)); do echo -n ; doneecho -ne ] $progress%\r echo $name # 获取开始时间model_start_time$(date %s) # 运行python脚本python gen_onnx.py $name ./log/$name 21python gen_rknn.py ./log/$name 21 # 记录结束时间并计算耗时model_end_time$(date %s)model_duration$((model_end_time - model_start_time)) output$(grep BUILD SUCCESS ./log/$name)echo $name $output (耗时: ${model_duration}s) # 增加已完成的模型计数count$((count 1))fi
done
# 打印新行以结束进度条
echo
EOF
bash run.sh
