fix(tools/scripts): missing `expanduser` (#1435)

* fix(tools/scripts): find env file failed (#1385)

* fix(tools/scripts): find env file failed

* Update quantize_model.md

* CI(ncnn): update ncnn version

docs/en/01-how-to-build/cross_build_ncnn_aarch64.md

@@ -64,7 +64,7 @@ ls -alh /tmp/ocv-aarch64
 c) Cross build ncnn and install to /tmp/ncnn-aarch64
 
 ```bash
-git clone https://github.com/tencent/ncnn --branch 20220729 --depth=1
+git clone https://github.com/tencent/ncnn --branch 20221128 --depth=1
 mkdir build && cd build
 cmake .. \
   -DCMAKE_TOOLCHAIN_FILE=../toolchains/aarch64-linux-gnu.toolchain.cmake \

docs/en/02-how-to-run/quantize_model.md

@@ -8,8 +8,8 @@ The fixed-point model has many advantages over the fp32 model:
 - Benefit from the smaller model, the Cache hit rate is improved and inference would be faster
 - Chips tend to have corresponding fixed-point acceleration instructions which are faster and less energy consumed (int8 on a common CPU requires only about 10% of energy)
 
-The size of the installation package and the heat generation are the key indicators of the mobile terminal evaluation APP;
-On the server side, quantization means that you can maintain the same QPS and improve model precision in exchange for improved accuracy.
+APK file size and heat generation are key indicators while evaluating mobile APP;
+On server side, quantization means that you can increase model size in exchange for precision and keep the same QPS.
 
 ## Post training quantization scheme
 
@@ -21,7 +21,7 @@ Taking ncnn backend as an example, the complete workflow is as follows:
 
 mmdeploy generates quantization table based on static graph (onnx) and uses backend tools to convert fp32 model to fixed point.
 
-Currently mmdeploy support ncnn with PTQ.
+mmdeploy currently supports ncnn with PTQ.
 
 ## How to convert model
 
@@ -38,10 +38,15 @@ Back in mmdeploy, enable quantization with the option 'tools/deploy.py --quant'.
 
 ```bash
 cd /path/to/mmdeploy
-export MODEL_PATH=/path/to/mmclassification/configs/resnet/resnet18_8xb16_cifar10.py
-export MODEL_CONFIG=https://download.openmmlab.com/mmclassification/v0/resnet/resnet18_b16x8_cifar10_20210528-bd6371c8.pth
 
-python3 tools/deploy.py  configs/mmcls/classification_ncnn-int8_static.py  ${MODEL_CONFIG}  ${MODEL_PATH}   /path/to/self-test.png   --work-dir work_dir --device cpu --quant --quant-image-dir /path/to/images
+export MODEL_CONFIG=/home/rg/konghuanjun/mmclassification/configs/resnet/resnet18_8xb32_in1k.py
+export MODEL_PATH=https://download.openmmlab.com/mmclassification/v0/resnet/resnet18_8xb32_in1k_20210831-fbbb1da6.pth
+
+# get some imagenet sample images
+git clone https://github.com/nihui/imagenet-sample-images --depth=1
+
+# quantize
+python3 tools/deploy.py  configs/mmcls/classification_ncnn-int8_static.py  ${MODEL_CONFIG}  ${MODEL_PATH}   /path/to/self-test.png   --work-dir work_dir --device cpu --quant --quant-image-dir /path/to/imagenet-sample-images
 ...
 ```
 
@@ -56,7 +61,7 @@ Description
 
 Calibration set is used to calculate quantization layer parameters. Some DFQ (Data Free Quantization) methods do not even require a dataset.
 
-- Create a new folder, just put in the picture (no directory structure required, no negative example required, no filename format required)
+- Create a folder, just put in some images (no directory structure, no negative example, no special filename format)
 - The image needs to be the data comes from real scenario otherwise the accuracy would be drop
 - You can not quantize model with test dataset
   | Type  | Train dataset | Validation dataset | Test dataset  | Calibration dataset |

docs/zh_cn/01-how-to-build/cross_build_ncnn_aarch64.md

@@ -64,7 +64,7 @@ ls -alh /tmp/ocv-aarch64
 c) 交叉编译 ncnn 安装到 tmp 目录
 
 ```bash
-git clone https://github.com/tencent/ncnn --branch 20220729 --depth=1
+git clone https://github.com/tencent/ncnn --branch 20221128 --depth=1
 mkdir build && cd build
 cmake .. \
   -DCMAKE_TOOLCHAIN_FILE=../toolchains/aarch64-linux-gnu.toolchain.cmake \

docs/zh_cn/02-how-to-run/quantize_model.md

@@ -40,7 +40,11 @@ cd /path/to/mmdeploy
 export MODEL_CONFIG=/path/to/mmclassification/configs/resnet/resnet18_8xb16_cifar10.py
 export MODEL_PATH=https://download.openmmlab.com/mmclassification/v0/resnet/resnet18_b16x8_cifar10_20210528-bd6371c8.pth
 
-python3 tools/deploy.py  configs/mmcls/classification_ncnn-int8_static.py  ${MODEL_CONFIG}  ${MODEL_PATH}   /path/to/self-test.png   --work-dir work_dir --device cpu --quant --quant-image-dir /path/to/images
+# 找一些 imagenet 样例图
+git clone https://github.com/nihui/imagenet-sample-images --depth=1
+
+# 量化模型
+python3 tools/deploy.py  configs/mmcls/classification_ncnn-int8_static.py  ${MODEL_CONFIG}  ${MODEL_PATH}   /path/to/self-test.png   --work-dir work_dir --device cpu --quant --quant-image-dir /path/to/imagenet-sample-images
 ...
 ```
 

tools/scripts/build_ubuntu_x64_ncnn.py

@@ -2,6 +2,7 @@
 import os
 import sys
 import time
+from pathlib import Path
 
 from ubuntu_utils import cmd_result, ensure_base_env, get_job
 
@@ -59,7 +60,7 @@ def install_pyncnn(dep_dir):
     # git clone
     if not os.path.exists('ncnn'):
         os.system(
-            'git clone --depth 1 --branch 20220729  https://github.com/tencent/ncnn && cd ncnn'  # noqa: E501
+            'git clone --depth 1 --branch 20221128  https://github.com/tencent/ncnn && cd ncnn'  # noqa: E501
         )
 
     ncnn_dir = os.path.join(dep_dir, 'ncnn')
@@ -175,7 +176,7 @@ def main():
     if install_mmdeploy(work_dir, dep_dir, ncnn_cmake_dir) != 0:
         return -1
 
-    if os.path.exists('~/mmdeploy.env'):
+    if os.path.exists(Path('~/mmdeploy.env').expanduser()):
         print('Please source ~/mmdeploy.env to setup your env !')
         os.system('cat ~/mmdeploy.env')
 

tools/scripts/build_ubuntu_x64_ort.py

@@ -2,6 +2,7 @@
 import os
 import sys
 import time
+from pathlib import Path
 
 from ubuntu_utils import ensure_base_env, get_job
 
@@ -94,7 +95,7 @@ def main():
     if install_mmdeploy(work_dir, ort_dir) != 0:
         return -1
 
-    if os.path.exists('~/mmdeploy.env'):
+    if os.path.exists(Path('~/mmdeploy.env').expanduser()):
         print('Please source ~/mmdeploy.env to setup your env !')
         os.system('cat ~/mmdeploy.env')
 

tools/scripts/build_ubuntu_x64_pplnn.py

@@ -2,6 +2,7 @@
 import os
 import sys
 import time
+from pathlib import Path
 
 from ubuntu_utils import cmd_result, ensure_base_env, get_job
 
@@ -150,7 +151,7 @@ def main():
                         build_cuda) != 0:
         return -1
 
-    if os.path.exists('~/mmdeploy.env'):
+    if os.path.exists(Path('~/mmdeploy.env').expanduser()):
         print('Please source ~/mmdeploy.env to setup your env !')
         os.system('cat ~/mmdeploy.env')
 

tools/scripts/build_ubuntu_x64_torchscript.py

@@ -2,6 +2,7 @@
 import os
 import sys
 import time
+from pathlib import Path
 
 from ubuntu_utils import (cmd_result, cu_version_name, ensure_base_env,
                           get_job, pytorch_version)
@@ -118,7 +119,7 @@ def main():
     if install_mmdeploy(work_dir, libtorch_dir) != 0:
         return -1
 
-    if os.path.exists('~/mmdeploy.env'):
+    if os.path.exists(Path('~/mmdeploy.env').expanduser()):
         print('Please source ~/mmdeploy.env to setup your env !')
         os.system('cat ~/mmdeploy.env')
 

tools/scripts/ubuntu_cross_build_aarch64.sh

@@ -33,7 +33,7 @@ build_ocv() {
   fi
   cd opencv/platforms/linux/cross_build_aarch64
   rm -rf CMakeCache.txt
-  cmake ../../.. -DCMAKE_INSTALL_PREFIX=/tmp/ocv-aarch64 -DCMAKE_TOOLCHAIN_FILE=../aarch64-gnu.toolchain.cmake
+  cmake ../../.. -DBUILD_TIFF=ON -DCMAKE_INSTALL_PREFIX=/tmp/ocv-aarch64 -DCMAKE_TOOLCHAIN_FILE=../aarch64-gnu.toolchain.cmake
   good_nproc
   jobs=$?
   make -j${jobs}
@@ -43,7 +43,7 @@ build_ocv() {
 
 build_ncnn() {
   if [ ! -e "ncnn" ];then
-    git clone https://github.com/tencent/ncnn --branch 20220729 --depth=1
+    git clone https://github.com/tencent/ncnn --branch 20221128 --depth=1
   fi
   if [ ! -e "ncnn/build_aarch64" ];then
     mkdir -p ncnn/build_aarch64

tools/scripts/ubuntu_cross_build_rknn.sh

@@ -60,7 +60,7 @@ build_ocv_arm_gnueabi() {
   cd opencv/build_arm_gnueabi
   rm -rf CMakeCache.txt
   cmake .. -DCMAKE_INSTALL_PREFIX=install -DCMAKE_TOOLCHAIN_FILE=../platforms/linux/arm-gnueabi.toolchain.cmake \
-    -DBUILD_PERF_TESTS=OFF -DBUILD_SHARED_LIBS=OFF -DBUILD_TESTS=OFF -DCMAKE_BUILD_TYPE=Release
+    -DBUILD_TIFF=ON -DBUILD_PERF_TESTS=OFF -DBUILD_SHARED_LIBS=OFF -DBUILD_TESTS=OFF -DCMAKE_BUILD_TYPE=Release
   good_nproc
   jobs=$?
   make -j${jobs} && make install