add pretrained model

README.md

@@ -23,6 +23,7 @@ Some key dependencies are listed below, while others are given in [`requirements
 - Python >= 3.6
 - PyTorch >= 1.3, and a corresponding version of torchvision
 - ffmpeg (used in data preparation)
+- Pre-trained models in the `pretrained` folder, which are listed in [`pretrained/README.md`](https://github.com/Siyu-C/ACAR-Net/blob/master/pretrained/README.md)
 
 ## Usage
 Default values for arguments `nproc_per_node`, `backend` and `master_port` are `8`, `nccl` and `31114` respectively.
@@ -36,7 +37,6 @@ In this case, the `master_addr` argument must be provided. Moreover, arguments `
 
 ## To-do List
 - Data preparation for AVA dataset
-- Pre-trained models
 - Model zoo
 - More advanced backbone
 - Data preparation for Kinetics dataset, and training on AVA-Kinetics

pretrained/README.md

@@ -0,0 +1,7 @@
+# Pre-trained Models
+
+Models are released in the Google drive folder [`pretrained`](https://drive.google.com/drive/folders/1enyXBbM47W0sqLACLau2m4DB_uDVWHke?usp=sharing), and should be downloaded to this local `pretrained` folder by default.
+
+Name | Architecture | Dataset | Source
+--- | :---: | :---: | :---:
+[`SLOWFAST_R50_K400.pth.tar`](https://drive.google.com/file/d/1kQO_dnM9JjV3sBtvowXqQa6d0xcxn5rs/view?usp=sharing) | SlowFast R-50 | Kinetics-400 | [PySlowFast](https://github.com/facebookresearch/slowfast)