# 计算 mAP
pip install pycocotools

!pip install mmcv-full==1.3.11 -f https://download.openmmlab.com/mmcv/dist/cu110/torch1.7.0/index.html 
!pip install openmim  > /dev/null
!mim install mmdet
!pip install fvcore>=0.1
!pip install pytorch-lightning>=1.3.8
!pip install timm>=0.4.12
!pip install pycocotools
!pip install pycuda
!pip install thop # 计算网络的flops

# apex 安装
git clone https://github.com/NVIDIA/apex.git
cd apex
pip install -v --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./
# pip install -v --no-cache-dir --global-option="--pyprof" --global-option="--cpp_ext" --global-option="--cuda_ext" ./
# python setup.py install --cpp_ext --cuda_ext

!pip install https://codechina.csdn.net/wc781708249/toolcv/archive/master.zip
# or
!pip install 'git+https://codechina.csdn.net/wc781708249/toolcv.git'
!pip install 'git+https://codechina.csdn.net/wc781708249/toolcv.git@master' # @切换分支
# or
!python setup.py install

!pip install 'git+https://codechina.csdn.net/wc781708249/toolcv.git' > /dev/null
!pip install mmcv-full==1.3.11 -f https://download.openmmlab.com/mmcv/dist/cu110/torch1.7.0/index.html  > /dev/null
!pip install openmim  > /dev/null
!mim install mmdet  > /dev/null

# pip install wandb

# Flexible integration for any Python script
import wandb

# 1. Start a W&B run
wandb.init(project='gpt3')

# 2. Save model inputs and hyperparameters
config = wandb.config
config.learning_rate = 0.01

# Model training here
‍
# 3. Log metrics over time to visualize performance
wandb.log({"loss": loss})