LSTM.py

#!/usr/bin/env python
# -*- coding: UTF-8 -*-
'''
@Project :Awesome-DL-Models 
@File    :LSTM.py
@Author  :JackHCC
@Date    :2022/3/14 19:04 
@Desc    :

'''
import torch
from torch import nn

from _utils import load_data_time_machine, train
from RNN import RNNModel, RNNModelScratch


def get_lstm_params(vocab_size, num_hiddens, device):
    num_inputs = num_outputs = vocab_size

    def normal(shape):
        return torch.randn(size=shape, device=device) * 0.01

    def three():
        return (normal((num_inputs, num_hiddens)),
                normal((num_hiddens, num_hiddens)),
                torch.zeros(num_hiddens, device=device))

    W_xi, W_hi, b_i = three()  # 输入门参数
    W_xf, W_hf, b_f = three()  # 遗忘门参数
    W_xo, W_ho, b_o = three()  # 输出门参数
    W_xc, W_hc, b_c = three()  # 候选记忆元参数
    # 输出层参数
    W_hq = normal((num_hiddens, num_outputs))
    b_q = torch.zeros(num_outputs, device=device)
    # 附加梯度
    params = [W_xi, W_hi, b_i, W_xf, W_hf, b_f, W_xo, W_ho, b_o, W_xc, W_hc,
              b_c, W_hq, b_q]
    for param in params:
        param.requires_grad_(True)
    return params


def init_lstm_state(batch_size, num_hiddens, device):
    return (torch.zeros((batch_size, num_hiddens), device=device),
            torch.zeros((batch_size, num_hiddens), device=device))


def lstm(inputs, state, params):
    [W_xi, W_hi, b_i, W_xf, W_hf, b_f, W_xo, W_ho, b_o, W_xc, W_hc, b_c,
     W_hq, b_q] = params
    (H, C) = state
    outputs = []
    for X in inputs:
        I = torch.sigmoid((X @ W_xi) + (H @ W_hi) + b_i)
        F = torch.sigmoid((X @ W_xf) + (H @ W_hf) + b_f)
        O = torch.sigmoid((X @ W_xo) + (H @ W_ho) + b_o)
        C_tilda = torch.tanh((X @ W_xc) + (H @ W_hc) + b_c)
        C = F * C + I * C_tilda
        H = O * torch.tanh(C)
        Y = (H @ W_hq) + b_q
        outputs.append(Y)
    return torch.cat(outputs, dim=0), (H, C)


if __name__ == "__main__":
    batch_size, num_steps = 32, 35
    train_iter, vocab = load_data_time_machine(batch_size, num_steps)

    vocab_size, num_hiddens, device = len(vocab), 256, torch.device('cuda' if torch.cuda.is_available() else 'cpu')
    num_epochs, lr = 500, 1

    num_inputs = vocab_size

    # LSTM输入的参数与RNN一致
    lstm_layer = nn.LSTM(num_inputs, num_hiddens)

    # Use API
    model = RNNModel(lstm_layer, len(vocab))
    model = model.to(device)

    # Do not use API
    model = RNNModelScratch(len(vocab), num_hiddens, device, get_lstm_params, init_lstm_state, lstm)

    train(model, train_iter, vocab, lr, num_epochs, device)