1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126 | import torch
import torch.nn as nn
from torch.nn import functional as F
# hyperparameters
batch_size = 32 # 独立的句子，并行计算数量
block_size = 8 # 上下文长度
max_iters = 3000 # 训练迭代次数
eval_interval = 300 # 评估间隔
learning_rate = 1e-2 # 学习率
device = 'cuda' if torch.cuda.is_available() else 'cpu'
eval_iters = 200 # 评估时的batch数
# ------------
torch.manual_seed(1337)
# wget https://raw.githubusercontent.com/karpathy/char-rnn/master/data/tinyshakespeare/input.txt
with open('input.txt', 'r', encoding='utf-8') as f:
text = f.read()
chars = sorted(list(set(text))) # 所有出现过的字符
vocab_size = len(chars) # 词表的长度 65
stoi = { ch:i for i,ch in enumerate(chars) } # 映射字典， key: char, value: idx
itos = { i:ch for i,ch in enumerate(chars) } # 映射字典， key: idx, value: char
encode = lambda s: [stoi[c] for c in s] # encoder: 给一个字符串，输出对应索引
decode = lambda l: ''.join([itos[i] for i in l]) # decoder: 给定索引，解码字符串
# 划分训练和验证集 （ 9：1）
data = torch.tensor(encode(text), dtype=torch.long)
n = int(0.9*len(data))
train_data = data[:n]
val_data = data[n:]
# 数据加载
def get_batch(split):
# generate a small batch of data of inputs x and targets y
data = train_data if split == 'train' else val_data
ix = torch.randint(len(data) - block_size, (batch_size,)) # 随机获取batch_size个起始字符索引
x = torch.stack([data[i:i+block_size] for i in ix]) # x输入字符序列
y = torch.stack([data[i+1:i+block_size+1] for i in ix]) # y：目标字符序列（右移一位）
x, y = x.to(device), y.to(device)
return x, y
@torch.no_grad()
def estimate_loss():
out = {}
model.eval()
for split in ['train', 'val']:
losses = torch.zeros(eval_iters)
for k in range(eval_iters):
X, Y = get_batch(split)
logits, loss = model(X, Y)
losses[k] = loss.item()
out[split] = losses.mean()
model.train()
return out
# super simple bigram model
class BigramLanguageModel(nn.Module):
def __init__(self, vocab_size):
super().__init__()
# each token directly reads off the logits for the next token from a lookup table
self.token_embedding_table = nn.Embedding(vocab_size, vocab_size)
def forward(self, idx, targets=None):
# idx and targets are both (B,T) tensor of integers
logits = self.token_embedding_table(idx) # (B,T,C)
if targets is None:
loss = None
else:
B, T, C = logits.shape
logits = logits.view(B*T, C)
targets = targets.view(B*T)
loss = F.cross_entropy(logits, targets)
return logits, loss
def generate(self, idx, max_new_tokens):
# idx：当前上下文字符序列（整数索引），形状 (B, T)
for _ in range(max_new_tokens):
# get the predictions, 调用forward
logits, loss = self(idx)
# 取最后一个时间步，因为 Bigram 模型只依赖当前最后一个字符
logits = logits[:, -1, :] # becomes (B, C)
# 转成概率分布
probs = F.softmax(logits, dim=-1) # (B, C)
# 根据概率采样下一个字符
idx_next = torch.multinomial(probs, num_samples=1) # (B, 1)
# 拼接到现有序列
idx = torch.cat((idx, idx_next), dim=1) # (B, T+1)
return idx # 形状 (B, T + max_new_tokens)
model = BigramLanguageModel(vocab_size)
m = model.to(device)
# create a PyTorch optimizer
optimizer = torch.optim.AdamW(model.parameters(), lr=learning_rate)
for iter in range(max_iters):
# 每 eval_interval = 300 步，评估训练集和验证集的平均 loss
# estimate_loss() 是前面定义的无梯度函数
if iter % eval_interval == 0:
losses = estimate_loss()
print(f"step {iter}: train loss {losses['train']:.4f}, val loss {losses['val']:.4f}")
# 获取一个batch的数据
xb, yb = get_batch('train')
# 前向 + loss
logits, loss = model(xb, yb)
# 清除上一步梯度
optimizer.zero_grad(set_to_none=True)
# 反向传播
loss.backward()
# 梯度更新
optimizer.step()
# generate from the model
context = torch.zeros((1, 1), dtype=torch.long, device=device)
print(decode(m.generate(context, max_new_tokens=500)[0].tolist()))
|