fansfi的博客

·在Colab上安装PyTorch/XLA·在TPU上运行基本的PyTorch函数，比如创建和添加张量。·在TPU上运行PyTorch模块和autograd。·在TPU上运行PyTorch网络。 在Colab TPU上使用张量、运行模块和运行整个网络非常简单运行的前提是：首先在主菜单上，单击Runtime并选择Change Runtime type。设置“TPU”为硬件加速器。下面的单元格确保您能够访问Colab上的TPU。选好TPUimport osassert os.environ['C

网上流传的几乎都是这种代码def see(): # cap = cv2.VideoCapture(path) cap = cv2.VideoCapture(0) mpHands = mp.solutions.hands hands = mpHands.Hands(static_image_mode=False, max_num_hands=2, min_detectio

一轮要跑几个小时import torchdevice = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")print(device)cpu原来是没用GPU点击修改 -> 笔记本设置->硬件加速器改为GPU，保存即可

import cv2import mediapipe as mpimport numpy as npcap = cv2.VideoCapture("./dataset/accident_3.mp4")# cap = cv2.VideoCapture("./Hands/video.mp4")# cap = cv2.VideoCapture(0)mpHands = mp.solutions.handshands = mpHands.Hands(static_image_mode=False,.

http://home.ustc.edu.cn/~hagjie/