Transformer Implementation

Implementation of transformer encoder-decoder model based on paper Attention is all you need.

It includes (all built from scratch on pyTorch):

• Encoder and Decoder layers
• Multi-Head Attention and Scaled Dot-Product Attention
• Add & Norm layer
• Position-wise Feed-Forward Networks
• Sinusoidal Positional Encodings
  All of them were implemented following paper..

Repository: https://github.com/rubzip/Transformers-Implementation