import math
import pandas as pd
import torch
from torch import nn
from d2l import torch as d2l

#@save
class PositionWiseFFN(nn.Module):
    """基于位置的前馈网络"""
    # 全连接
    # num_step会变 序列长度
    # 所以序列当中每一个元素做一个全连接
    # (batch_size, num_step, ffn_num_input) ->  (batch_size, num_step, ffn_num_outputs)
    def __init__(self, ffn_num_input, ffn_num_hiddens, ffn_num_outputs,
                 **kwargs):
        super(PositionWiseFFN, self).__init__(**kwargs)
        self.dense1 = nn.Linear(ffn_num_input, ffn_num_hiddens)
        self.relu = nn.ReLU()
        self.dense2 = nn.Linear(ffn_num_hiddens, ffn_num_outputs)

    def forward(self, X):
        # 当输入不是二维时，前面的都当作样本维，最后一维当作特征维
        return self.dense2(self.relu(self.dense1(X)))
# 改变张量的最里层维度的尺寸，会改变成基于位置的前馈网络的输出尺寸。
# 单隐藏层的MLP
ffn = PositionWiseFFN(4, 4, 8)
ffn.eval()
print('ffn.shape : ', ffn(torch.ones((2, 3, 4))).shape)
"""ffn.shape :  torch.Size([2, 3, 8])"""

# 层归一化
# 样本 比较稳定，不会应为长度发生变化而变化太大
ln = nn.LayerNorm(3)
# 特征 长度不同
bn = nn.BatchNorm1d(3)
X = torch.tensor([[2, 3, 1], [3, 1, 2]], dtype=torch.float32)
# X = torch.arange(18, dtype=torch.float32).reshape(2, 3, 3)
# 在训练模式下计算X的均值和方差
print('layer norm:', ln(X), '\nbatch norm:', bn(X))
"""
layer norm: tensor([[ 0.0000,  1.2247, -1.2247],
        [ 1.2247, -1.2247,  0.0000]], grad_fn=<NativeLayerNormBackward0>) 
batch norm: tensor([[-1.0000,  1.0000, -1.0000],
        [ 1.0000, -1.0000,  1.0000]], grad_fn=<NativeBatchNormBackward0>)
"""
# 残差连接和层规范化
#@save
class AddNorm(nn.Module):
    """残差连接后进行层规范化"""
    def __init__(self, normalized_shape, dropout, **kwargs):
        super(AddNorm, self).__init__(**kwargs)
        self.dropout = nn.Dropout(dropout)
        # 输入为(1, 3, 5, 5)，layernorm的normalized_shape为[3, 5, 5]，也就是说对后三维度进行归一化操作
        self.ln = nn.LayerNorm(normalized_shape)

    def forward(self, X, Y):
        return self.ln(self.dropout(Y) + X)

add_norm = AddNorm([3, 4], 0.5)
add_norm.eval()
print('add_norm.shape : ', add_norm(torch.ones((2, 3, 4)), torch.ones((2, 3, 4))).shape)
"""add_norm.shape :  torch.Size([2, 3, 4])"""
# 实现编码器中的一个层
#@save
class EncoderBlock(nn.Module):
    """transformer编码器块"""
    def __init__(self, key_size, query_size, value_size, num_hiddens,
                 norm_shape, ffn_num_input, ffn_num_hiddens, num_heads,
                 dropout, use_bias=False, **kwargs):
        super(EncoderBlock, self).__init__(**kwargs)
        self.attention = d2l.MultiHeadAttention(
            key_size, query_size, value_size, num_hiddens, num_heads, dropout,
            use_bias)
        self.addnorm1 = AddNorm(norm_shape, dropout)
        self.ffn = PositionWiseFFN(ffn_num_input, ffn_num_hiddens, num_hiddens)
        self.addnorm2 = AddNorm(norm_shape, dropout)

    def forward(self, X, valid_lens):
        Y = self.addnorm1(X, self.attention(X, X, X, valid_lens))
        return self.addnorm2(Y, self.ffn(Y))

# transformer编码器中的任何层都不会改变其输入的形状
X = torch.ones((2, 100, 24))
valid_lens = torch.tensor([3, 2])
# (key_size, query_size, value_size, num_hiddens,norm_shape,
#  ffn_num_input, ffn_num_hiddens, num_heads,dropout)
encoder_blk = EncoderBlock(24, 24, 24, 24, [100, 24], 24, 48, 8, 0.5)
print('encoder.shape : ', encoder_blk(X, valid_lens).shape)
"""encoder.shape :  torch.Size([2, 100, 24])"""
# transformer编码器
#@save
class TransformerEncoder(d2l.Encoder):
    """transformer编码器"""
    def __init__(self, vocab_size, key_size, query_size, value_size,
                 num_hiddens, norm_shape, ffn_num_input, ffn_num_hiddens,
                 num_heads, num_layers, dropout, use_bias=False, **kwargs):
        super(TransformerEncoder, self).__init__(**kwargs)
        self.num_hiddens = num_hiddens
        self.embedding = nn.Embedding(vocab_size, num_hiddens)
        self.pos_encoding = d2l.PositionalEncoding(num_hiddens, dropout)
        self.blks = nn.Sequential()
        for i in range(num_layers):
            self.blks.add_module(
                "block"+str(i),
                EncoderBlock(key_size, query_size, value_size, num_hiddens,
                    norm_shape, ffn_num_input, ffn_num_hiddens,
                    num_heads, dropout, use_bias)
            )

    def forward(self, X, valid_lens, *args):
        # 因为位置编码值在-1和1之间，
        # 因此嵌入值乘以嵌入维度的平方根进行缩放，然后再与位置编码相加。
        # 大小要匹配
        X = self.pos_encoding(self.embedding(X) * math.sqrt(self.num_hiddens))
        # 权重图
        self.attention_weights = [None] * len(self.blks)
        # enumerate多用于在for循环中得到计数，利用它可以同时获得索引和值，
        # 即需要index和value值的时候可以使用enumerate
        for i, blk in enumerate(self.blks):
            X = blk(X, valid_lens)
            self.attention_weights[i] = blk.attention.attention.attention_weights
        return X

encoder = TransformerEncoder(200, 24, 24, 24, 24, [100, 24], 24, 48, 8, 2, 0.5)
encoder.eval()
print('encoder.shape : ', encoder(torch.ones((2, 100), dtype=torch.long), valid_lens).shape)
"""encoder.shape :  torch.Size([2, 100, 24])"""
# transformer解码器也是由多个相同的层组成
class DecoderBlock(nn.Module):
    """解码器中第i个块"""
    def __init__(self, key_size, query_size, value_size, num_hiddens,
                 norm_shape, ffn_num_input, ffn_num_hiddens, num_heads,
                 dropout, i, **kwargs):
        super(DecoderBlock, self).__init__(**kwargs)
        self.i = i
        self.attention1 = d2l.MultiHeadAttention(
            key_size, query_size, value_size, num_hiddens, num_heads, dropout)
        self.addnorm1 = AddNorm(norm_shape, dropout)
        self.attention2 = d2l.MultiHeadAttention(
            key_size, query_size, value_size, num_hiddens, num_heads, dropout)
        self.addnorm2 = AddNorm(norm_shape, dropout)
        self.ffn = PositionWiseFFN(ffn_num_input, ffn_num_hiddens, num_hiddens)
        self.addnorm3 = AddNorm(norm_shape, dropout)

    def forward(self, X, state):
        enc_outputs, enc_valid_lens = state[0], state[1]
        # 训练阶段，输出序列的所有词元都在同一时间处理，
        # 因此state[2][self.i]初始化为None。
        # 预测阶段，输出序列是通过词元一个接着一个解码的，
        # 因此state[2][self.i]包含着直到当前时间步第i个块解码的输出表示
        if state[2][self.i] is None:
            key_values = X
        else:
            key_values = torch.cat((state[2][self.i], X), axis=1)
        state[2][self.i] = key_values
        if self.training:
            batch_size, num_steps, _ = X.shape
            # dec_valid_lens的开头:(batch_size,num_steps),
            # 其中每一行是[1,2,...,num_steps]
            dec_valid_lens = torch.arange(
                1, num_steps + 1, device=X.device).repeat(batch_size, 1)
        else:
            dec_valid_lens = None

        # 自注意力
        # dec_valid_lens training时不关注之后的内容
        X2 = self.attention1(X, key_values, key_values, dec_valid_lens)
        Y = self.addnorm1(X, X2)
        # 编码器－解码器注意力。
        # enc_outputs的开头:(batch_size,num_steps,num_hiddens)
        # enc_valid_lens 哪些为padding的内容
        Y2 = self.attention2(Y, enc_outputs, enc_outputs, enc_valid_lens)
        Z = self.addnorm2(Y, Y2)
        return self.addnorm3(Z, self.ffn(Z)), state

decoder_blk = DecoderBlock(24, 24, 24, 24, [100, 24], 24, 48, 8, 0.5, 0)
decoder_blk.eval()
X = torch.ones((2, 100, 24))
state = [encoder_blk(X, valid_lens), valid_lens, [None]]
print('decoder.shape : ', decoder_blk(X, state)[0].shape)


# transformer解码器
class TransformerDecoder(d2l.AttentionDecoder):
    def __init__(self, vocab_size, key_size, query_size, value_size,
                 num_hiddens, norm_shape, ffn_num_input, ffn_num_hiddens,
                 num_heads, num_layers, dropout, **kwargs):
        super(TransformerDecoder, self).__init__(**kwargs)
        self.num_hiddens = num_hiddens
        self.num_layers = num_layers
        self.embedding = nn.Embedding(vocab_size, num_hiddens)
        self.pos_encoding = d2l.PositionalEncoding(num_hiddens, dropout)
        self.blks = nn.Sequential()
        for i in range(num_layers):
            self.blks.add_module(
                "block"+str(i),
                DecoderBlock(key_size, query_size, value_size, num_hiddens,
                             norm_shape, ffn_num_input, ffn_num_hiddens,
                             num_heads, dropout, i))
        self.dense = nn.Linear(num_hiddens, vocab_size)

    def init_state(self, enc_outputs, enc_valid_lens, *args):
        return [enc_outputs, enc_valid_lens, [None] * self.num_layers]

    def forward(self, X, state):
        # 因为位置编码值在-1和1之间，
        # 因此嵌入值乘以嵌入维度的平方根进行缩放，然后再与位置编码相加。
        # 大小要匹配
        X = self.pos_encoding(self.embedding(X) * math.sqrt(self.num_hiddens))
        # 注意力权重
        self._attention_weights = [[None] * len(self.blks) for _ in range (2)]
        for i, blk in enumerate(self.blks):
            X, state = blk(X, state)
            # 解码器自注意力权重
            self._attention_weights[0][i] = blk.attention1.attention.attention_weights
            # “编码器－解码器”自注意力权重
            self._attention_weights[1][i] = blk.attention2.attention.attention_weights
        return self.dense(X), state

    @property
    def attention_weights(self):
        return self._attention_weights

# 训练
num_hiddens, num_layers, dropout, batch_size, num_steps = 32, 2, 0.1, 64, 10
lr, num_epochs, device = 0.005, 200, d2l.try_gpu()
ffn_num_input, ffn_num_hiddens, num_heads = 32, 64, 4
key_size, query_size, value_size = 32, 32, 32
norm_shape = [32]

train_iter, src_vocab, tgt_vocab = d2l.load_data_nmt(batch_size, num_steps)

encoder = TransformerEncoder(
    len(src_vocab), key_size, query_size, value_size, num_hiddens,
    norm_shape, ffn_num_input, ffn_num_hiddens, num_heads,
    num_layers, dropout)
decoder = TransformerDecoder(
    len(tgt_vocab), key_size, query_size, value_size, num_hiddens,
    norm_shape, ffn_num_input, ffn_num_hiddens, num_heads,
    num_layers, dropout)
net = d2l.EncoderDecoder(encoder, decoder)
d2l.train_seq2seq(net, train_iter, lr, num_epochs, tgt_vocab, device)

engs = ['go .', "i lost .", 'he\'s calm .', 'i\'m home .']
fras = ['va !', 'j\'ai perdu .', 'il est calme .', 'je suis chez moi .']
for eng, fra in zip(engs, fras):
    translation, dec_attention_weight_seq = d2l.predict_seq2seq(
        net, eng, src_vocab, tgt_vocab, num_steps, device, True)
    print(f'{eng} => {translation}, ',
          f'bleu {d2l.bleu(translation, fra, k=2):.3f}')