免费试用Kaggle的GPU,进行深度学习，每周可以使用41小时GPU和20小时TPU

进入官网：Kaggle: Your Home for Data Science

然后注册一个账号，创建一个Notebook

设置为GPU

 

Adding a free GPU

You can add a single NVIDIA Tesla P100 to your Notebook for free. GPU environments have lower CPU and main memory, but are a great way to achieve significant speed-ups for certain types of work like training neural networks on image data. One of the major benefits to using Notebooks as opposed to a local machine or your own VM is that the Notebook environment is already pre-configured with GPU-ready software and packages which can be time consuming and frustrating to set-up. Free GPU availability is limited: in busy times, you might be placed in a queue.

To add a GPU, navigate to the “Settings” pane from the Notebook editor and click the “Accelerator" > GPU option. Your session will restart which may take a few moments to several minutes if you don’t need to wait in a queue to access a GPU-enabled machine.

To learn more about getting the most out of using a GPU in Notebooks, check out this tutorial Notebook by Dan Becker.

Adding a free TPU

You can add a TPU v3-8 to your Notebook for free. TPUs are hardware accelerators specialized in deep learning tasks. They are supported in Tensorflow 2.1 both through the Keras high-level API and, at a lower level, in models using a custom training loop. Free TPU availability is limited: in busy times, you might be placed in a queue. To learn more about getting the most out of using a TPU in Notebooks, check out this in depth guide.

To add a TPU, navigate to the “Settings” pane from the Notebook editor and click the “Accelerator" > TPU v3-8 option. Your session will restart which may take a few moments to several minutes if you don’t need to wait in a queue to access a TPU-enabled machine.