Attention Is All You Need [1]

[1]

1. Encoder-Decoder

• The encoder maps an input sequence of symbol representations to a sequence of continuous representations .
• Given , the decoder then generates an output sequence of symbols one element at a time.
• At each step the model is auto-regressive, consuming the previously generated symbols as additional input when generating the next.

Overview of Transformer

• Encoder
  • It has 6 identical layers.
  • Each layer has a multi-head self-attention sub-layer and a position-wise fully connect feed-forward sub-layer in turn.
  • Each sub-layer uses the residual connection mechanism, and followed by a normalization layer.
  • The residual connection mechanism uses x+F(x) as its final result.
• Decoder
  • It has 6 identical layers.
  • Each layer has three sub-layers:
    • Masked multi-head attention layer ensues the prediction at time t can only depends on the known output at positions less than t.
    • Multi-head attention layer further adds the output of the encoder stack into this decoder.
    • Position-wise fully connect feed-forward layer.

2. Attention

• Mapping a query and a set of key-value pairs to a weighted output.
• Scaled Dot-product Attention
  
  Scaled Dot-product Attention
  • Given three vectors, a query , keys and values :
    
• Multi-head Attention
  
  Multi-head Attention
  • Jointly collect information from different representation subspace focused on different positions.
  • Given queries , keys and values :
    
    
    • , , , , , .
    • The and have the same dimension .
    • First, it linearly projects the queries, keys and values times to learn different , and .
    • Next, it concatenates all yield output values together.
    • At last, it projects the concatenated vector into a -dimension vector.
      
      Example [2]
• Attention in Transformer
  • Encoder's multi-head:
    • ===the output of previous layer.
    • Each position in the encoder can attend to all positions in the previous layer of the encoder.
  • Decoder's masked multi-head:
    • ===the masked output of previous layer.
    • For example, if we predict the -th output token, all tokens after timestamp have to be marked.
    • This prevents leftward information flow in the decoder in order to preserve the auto-regressive property.
    • It masks out (setting to ) all values in the input of the softmax which correspond to illegal connections during the scaled dot-product attention.
  • Decoder's multi-head:
    • =the output of the previous decoder layer, ==the encoder stack's output.
    • This allows every position in the decoder to attend over all positions in the input sequence.

3. Position-wise Feed-forward Networks

• ReLu activation function: .
• It's applied to each position separately and identically.

4. Positional Encoding

• To make use of the order of the sequence.
• Add at the bottoms of the encoder and decoder stacks.
• Have the same dimension as the embeddings.
• 
  
  • is the index of position, in encoder while in decoder.
  • is the index of the dimension , .

Reference

[1] Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., ... & Polosukhin, I. (2017). Attention is all you need. In Advances in neural information processing systems (pp. 5998-6008).
[2] 口仆. Transformer 原理解析 https://zhuanlan.zhihu.com/p/135873679