TransformerLayer.py

'''
This is our own transformer layer implementation

https://neptune.ai/blog/how-to-code-bert-using-pytorch-tutorial
code from :https://github.com/codertimo/BERT-pytorch/tree/d10dc4f9d5a6f2ca74380f62039526eb7277c671/bert_pytorch

'''

import torch
import torch.nn as nn
import math
import torch.nn.functional as F


class LayerNorm(nn.Module):
    "Construct a layernorm module (See citation for details)."

    def __init__(self, features, eps=1e-6):
        super(LayerNorm, self).__init__()
        self.a_2 = nn.Parameter(torch.ones(features))
        self.b_2 = nn.Parameter(torch.zeros(features))
        self.eps = eps

    def forward(self, x):
        mean = x.mean(-1, keepdim=True)
        std = x.std(-1, keepdim=True)
        return self.a_2 * (x - mean) / (std + self.eps) + self.b_2

class SublayerConnection(nn.Module):
    """
    A residual connection followed by a layer norm.
    Note for code simplicity the norm is first as opposed to last.
    """

    def __init__(self, size, dropout):
        super(SublayerConnection, self).__init__()
        self.norm = LayerNorm(size)
        self.dropout = nn.Dropout(dropout)

    def forward(self, x, sublayer):
        "Apply residual connection to any sublayer with the same size."
        return x + self.dropout(sublayer(self.norm(x)))

class PositionwiseFeedForward(nn.Module):
    "Implements FFN equation."

    def __init__(self, d_model, d_ff, dropout=0.1):
        super(PositionwiseFeedForward, self).__init__()
        self.w_1 = nn.Linear(d_model, d_ff)
        self.w_2 = nn.Linear(d_ff, d_model)
        self.dropout = nn.Dropout(dropout)
        self.activation = GELU()

    def forward(self, x):
        return self.w_2(self.dropout(self.activation(self.w_1(x))))

class GELU(nn.Module):
    """
    Paper Section 3.4, last paragraph notice that BERT used the GELU instead of RELU
    """
    def forward(self, x):
        return 0.5 * x * (1 + torch.tanh(math.sqrt(2 / math.pi) * (x + 0.044715 * torch.pow(x, 3))))

class TokenEmbedding(nn.Embedding):
    def __init__(self, vocab_size, embed_size=512):
        super(TokenEmbedding,self).__init__(vocab_size, embed_size, padding_idx=0)

class SegmentEmbedding(nn.Embedding):
    def __init__(self, embed_size=512):
        super(SegmentEmbedding,self).__init__(3, embed_size, padding_idx=0)

class PositionalEmbedding(nn.Module):

    def __init__(self, d_model, max_len=512):
        super(PositionalEmbedding,self).__init__()

        # Compute the positional encodings once in log space.
        pe = torch.zeros(max_len, d_model).float()
        pe.require_grad = False

        position = torch.arange(0, max_len).float().unsqueeze(1)
        div_term = (torch.arange(0, d_model, 2).float() * -(math.log(10000.0) / d_model)).exp()

        pe[:, 0::2] = torch.sin(position * div_term)
        pe[:, 1::2] = torch.cos(position * div_term)

        pe = pe.unsqueeze(0)
        self.register_buffer('pe', pe)

    def forward(self, x):
        return self.pe[:, :x.size(1)]



class BERTEmbedding(nn.Module):
    """
    BERT Embedding which is consisted with under features
        1. TokenEmbedding : normal embedding matrix
        2. PositionalEmbedding : adding positional information using sin, cos
        2. SegmentEmbedding : adding sentence segment info, (sent_A:1, sent_B:2)
        sum of all these features are output of BERTEmbedding
    """

    def __init__(self, vocab_size, embed_size, dropout=0.1):
        """
        :param vocab_size: total vocab size
        :param embed_size: embedding size of token embedding
        :param dropout: dropout rate
        """
        super(BERTEmbedding,self).__init__()
        self.token = TokenEmbedding(vocab_size=vocab_size, embed_size=embed_size)
        self.position = PositionalEmbedding(d_model=self.token.embedding_dim)
        self.segment = SegmentEmbedding(embed_size=self.token.embedding_dim)
        self.dropout = nn.Dropout(p=dropout)
        self.embed_size = embed_size

    def forward(self, sequence, segment_label):
        x = self.token(sequence) + self.position(sequence) + self.segment(segment_label)
        return self.dropout(x)

class Attention(nn.Module):
    """
    Compute 'Scaled Dot Product Attention
    """

    def forward(self, query, key, value, mask=None, dropout=None):
        scores = torch.matmul(query, key.transpose(-2, -1)) \
                 / math.sqrt(query.size(-1))

        if mask is not None:
            scores = scores.masked_fill(mask == 0, -1e9)

        p_attn = F.softmax(scores, dim=-1)

        if dropout is not None:
            p_attn = dropout(p_attn)

        return torch.matmul(p_attn, value), p_attn

class MultiHeadedAttention(nn.Module):
    """
    Take in model size and number of heads.
    """

    def __init__(self, h, d_model, dropout=0.1):
        super(MultiHeadedAttention,self).__init__()
        assert d_model % h == 0

        # We assume d_v always equals d_k
        self.d_k = d_model // h
        self.h = h

        self.linear_layers = nn.ModuleList([nn.Linear(d_model, d_model) for _ in range(3)])
        self.output_linear = nn.Linear(d_model, d_model)
        self.attention = Attention()

        self.dropout = nn.Dropout(p=dropout)

    def forward(self, query, key, value, mask=None):
        batch_size = query.size(0)

        # 1) Do all the linear projections in batch from d_model => h x d_k
        query, key, value = [l(x).view(batch_size, -1, self.h, self.d_k).transpose(1, 2)
                             for l, x in zip(self.linear_layers, (query, key, value))]

        # 2) Apply attention on all the projected vectors in batch.
        x, attn = self.attention(query, key, value, mask=mask, dropout=self.dropout)

        # 3) "Concat" using a view and apply a final linear.
        x = x.transpose(1, 2).contiguous().view(batch_size, -1, self.h * self.d_k)

        return self.output_linear(x)


class TransformerBlock(nn.Module):
    """
    Bidirectional Encoder = Transformer (self-attention)
    Transformer = MultiHead_Attention + Feed_Forward with sublayer connection
    """

    def __init__(self, hidden, attn_heads, feed_forward_hidden, dropout):
        """
        :param hidden: hidden size of transformer
        :param attn_heads: head sizes of multi-head attention
        :param feed_forward_hidden: feed_forward_hidden, usually 4*hidden_size
        :param dropout: dropout rate
        """

        super(TransformerBlock,self).__init__()
        self.attention = MultiHeadedAttention(h=attn_heads, d_model=hidden)
        self.feed_forward = PositionwiseFeedForward(d_model=hidden, d_ff=feed_forward_hidden, dropout=dropout)
        self.input_sublayer = SublayerConnection(size=hidden, dropout=dropout)
        self.output_sublayer = SublayerConnection(size=hidden, dropout=dropout)
        self.dropout = nn.Dropout(p=dropout)

    def forward(self, x):


        x = self.input_sublayer(x, lambda _x: self.attention.forward(_x, _x, _x, mask=None))


        x = self.output_sublayer(x, self.feed_forward)



        return self.dropout(x)


class BERT(nn.Module):
    """
    BERT model : Bidirectional Encoder Representations from Transformers.
    """

    def __init__(self, hidden=768, n_layers=12, attn_heads=12, dropout=0.1):
        """
        :param vocab_size: vocab_size of total words
        :param hidden: BERT model hidden size
        :param n_layers: numbers of Transformer blocks(layers)
        :param attn_heads: number of attention heads
        :param dropout: dropout rate
        """

        super(BERT, self).__init__()
        self.hidden = hidden
        self.n_layers = n_layers
        self.attn_heads = attn_heads

        # paper noted they used 4*hidden_size for ff_network_hidden_size
        self.feed_forward_hidden = hidden * 4

        # embedding for BERT, sum of positional, segment, token embeddings
        # self.embedding = BERTEmbedding(vocab_size=vocab_size, embed_size=hidden)

        # multi-layers transformer blocks, deep network
        self.transformer_blocks = nn.ModuleList(
            [TransformerBlock(hidden, attn_heads, hidden * 4, dropout) for _ in range(n_layers)])

    def forward(self, x):
        # attention masking for padded token
        # torch.ByteTensor([batch_size, 1, seq_len, seq_len)
        # mask = (x > 0).unsqueeze(1).repeat(1, x.size(1), 1).unsqueeze(1)

        # embedding the indexed sequence to sequence of vectors
        # x = self.embedding(x, segment_info)

        # running over multiple transformer blocks
        for transformer in self.transformer_blocks:
            x = transformer.forward(x)



        'same shape'
        # torch.Size([2, 3, 128])

        return x


import argparse



def train():
    parser = argparse.ArgumentParser()

    # parser.add_argument("-c", "--train_dataset", required=True, type=str, help="train dataset for train bert")
    # parser.add_argument("-t", "--test_dataset", type=str, default=None, help="test set for evaluate train set")
    # parser.add_argument("-v", "--vocab_path", required=True, type=str, help="built vocab model path with bert-vocab")
    # parser.add_argument("-o", "--output_path", required=True, type=str, help="ex)output/bert.model")

    parser.add_argument("-hs", "--hidden", type=int, default=128, help="hidden size of transformer model")
    parser.add_argument("-l", "--layers", type=int, default=2, help="number of layers")
    parser.add_argument("-a", "--attn_heads", type=int, default=8, help="number of attention heads")
    parser.add_argument("-s", "--seq_len", type=int, default=20, help="maximum sequence len")

    parser.add_argument("-b", "--batch_size", type=int, default=64, help="number of batch_size")
    parser.add_argument("-e", "--epochs", type=int, default=10, help="number of epochs")
    parser.add_argument("-w", "--num_workers", type=int, default=5, help="dataloader worker size")

    parser.add_argument("--with_cuda", type=bool, default=True, help="training with CUDA: true, or false")
    parser.add_argument("--log_freq", type=int, default=10, help="printing loss every n iter: setting n")
    parser.add_argument("--corpus_lines", type=int, default=None, help="total number of lines in corpus")
    parser.add_argument("--cuda_devices", type=int, nargs='+', default=None, help="CUDA device ids")
    parser.add_argument("--on_memory", type=bool, default=True, help="Loading on memory: true or false")

    parser.add_argument("--lr", type=float, default=1e-3, help="learning rate of adam")
    parser.add_argument("--adam_weight_decay", type=float, default=0.01, help="weight_decay of adam")
    parser.add_argument("--adam_beta1", type=float, default=0.9, help="adam first beta value")
    parser.add_argument("--adam_beta2", type=float, default=0.999, help="adam first beta value")

    args = parser.parse_args()

    device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
    print('Using device:', device)

    print("Building BERT model")
    bert = BERT(hidden=args.hidden, n_layers=args.layers, attn_heads=args.attn_heads).to(device)


    test = torch.rand([2, 3, 128]).to(device)

    output = bert(test)
    import pdb;
    pdb.set_trace()

    print ('BERT model...', bert)


if __name__ == "__main__":
    train()