EasyCV/docs

maintain docs

1. install requirements needed to build docs

   # in easycv root dir
   pip install requirements/docs.txt
   

2. build docs

   # in easycv/docs dir
   bash build_docs.sh
   

3. doc string format

   We adopt the google style docstring format as the standard, please refer to the following documents.

   1. Google Python style guide docstring link
   2. Google docstring example link
   3. sample：torch.nn.modules.conv link
   4. Transformer as an example：

   class Transformer(base.Layer):
       """
           Transformer model from ``Attention Is All You Need``,
           Original paper: https://arxiv.org/abs/1706.03762
   
           Args:
               num_token (int): vocab size.
               num_layer (int): num of layer.
               num_head (int): num of attention heads.
               embedding_dim (int): embedding dimension.
               attention_head_dim (int): attention head dimension.
               feed_forward_dim (int): feed forward dimension.
               initializer: initializer type.
               activation: activation function.
               dropout (float): dropout rate (0.0 to 1.0).
               attention_dropout (float): dropout rate for attention layer.
   
           Returns: None
       """