Yeqing Li
Yeqing Li
For the CoAtNet, the priority changes. We current do not have dedicated resources to reimplement the CoAtNet model. We will revisit the feature request if this became more popular request...
The problem for the Kinetics-400 pretrain is that you need to download the data yourself and format them into the tfrecrod files. You can refer to [here](https://github.com/tensorflow/models/blob/473f34aa986af7aba63509f4e25be1269f084619/official/vision/dataloaders/video_input_test.py#L31-L57) for the required...
Please submit customized script that is used for your own project to your specific repo. The script include hardcoded file names, which is not the common function needed for this...