-
Notifications
You must be signed in to change notification settings - Fork 6.4k
/
Copy pathch04.py
217 lines (173 loc) · 7.18 KB
/
ch04.py
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
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
# Copyright (c) Sebastian Raschka under Apache License 2.0 (see LICENSE.txt).
# Source for "Build a Large Language Model From Scratch"
# - https://www.manning.com/books/build-a-large-language-model-from-scratch
# Code: https://github.com/rasbt/LLMs-from-scratch
from .ch03 import MultiHeadAttention, PyTorchMultiHeadAttention
import torch
import torch.nn as nn
class LayerNorm(nn.Module):
def __init__(self, emb_dim):
super().__init__()
self.eps = 1e-5
self.scale = nn.Parameter(torch.ones(emb_dim))
self.shift = nn.Parameter(torch.zeros(emb_dim))
def forward(self, x):
mean = x.mean(dim=-1, keepdim=True)
var = x.var(dim=-1, keepdim=True, unbiased=False)
norm_x = (x - mean) / torch.sqrt(var + self.eps)
return self.scale * norm_x + self.shift
class GELU(nn.Module):
def __init__(self):
super().__init__()
def forward(self, x):
return 0.5 * x * (1 + torch.tanh(
torch.sqrt(torch.tensor(2.0 / torch.pi)) *
(x + 0.044715 * torch.pow(x, 3))
))
class FeedForward(nn.Module):
def __init__(self, cfg):
super().__init__()
self.layers = nn.Sequential(
nn.Linear(cfg["emb_dim"], 4 * cfg["emb_dim"]),
GELU(),
nn.Linear(4 * cfg["emb_dim"], cfg["emb_dim"]),
)
def forward(self, x):
return self.layers(x)
class TransformerBlock(nn.Module):
def __init__(self, cfg):
super().__init__()
self.att = MultiHeadAttention(
d_in=cfg["emb_dim"],
d_out=cfg["emb_dim"],
context_length=cfg["context_length"],
num_heads=cfg["n_heads"],
dropout=cfg["drop_rate"],
qkv_bias=cfg["qkv_bias"])
self.ff = FeedForward(cfg)
self.norm1 = LayerNorm(cfg["emb_dim"])
self.norm2 = LayerNorm(cfg["emb_dim"])
self.drop_resid = nn.Dropout(cfg["drop_rate"])
def forward(self, x):
# Shortcut connection for attention block
shortcut = x
x = self.norm1(x)
x = self.att(x) # Shape [batch_size, num_tokens, emb_size]
x = self.drop_resid(x)
x = x + shortcut # Add the original input back
# Shortcut connection for feed-forward block
shortcut = x
x = self.norm2(x)
x = self.ff(x)
x = self.drop_resid(x)
x = x + shortcut # Add the original input back
return x
class GPTModel(nn.Module):
def __init__(self, cfg):
super().__init__()
self.tok_emb = nn.Embedding(cfg["vocab_size"], cfg["emb_dim"])
self.pos_emb = nn.Embedding(cfg["context_length"], cfg["emb_dim"])
self.drop_emb = nn.Dropout(cfg["drop_rate"])
self.trf_blocks = nn.Sequential(
*[TransformerBlock(cfg) for _ in range(cfg["n_layers"])])
self.final_norm = LayerNorm(cfg["emb_dim"])
self.out_head = nn.Linear(cfg["emb_dim"], cfg["vocab_size"], bias=False)
def forward(self, in_idx):
batch_size, seq_len = in_idx.shape
tok_embeds = self.tok_emb(in_idx)
pos_embeds = self.pos_emb(torch.arange(seq_len, device=in_idx.device))
x = tok_embeds + pos_embeds # Shape [batch_size, num_tokens, emb_size]
x = self.drop_emb(x)
x = self.trf_blocks(x)
x = self.final_norm(x)
logits = self.out_head(x)
return logits
def generate_text_simple(model, idx, max_new_tokens, context_size):
# idx is (B, T) array of indices in the current context
for _ in range(max_new_tokens):
# Crop current context if it exceeds the supported context size
# E.g., if LLM supports only 5 tokens, and the context size is 10
# then only the last 5 tokens are used as context
idx_cond = idx[:, -context_size:]
# Get the predictions
with torch.no_grad():
logits = model(idx_cond)
# Focus only on the last time step
# (batch, n_token, vocab_size) becomes (batch, vocab_size)
logits = logits[:, -1, :]
# Get the idx of the vocab entry with the highest logits value
idx_next = torch.argmax(logits, dim=-1, keepdim=True) # (batch, 1)
# Append sampled index to the running sequence
idx = torch.cat((idx, idx_next), dim=1) # (batch, n_tokens+1)
return idx
######################
# Bonus
######################
class FeedForwardFast(nn.Module):
def __init__(self, cfg):
super().__init__()
self.layers = nn.Sequential(
nn.Linear(cfg["emb_dim"], 4 * cfg["emb_dim"]),
nn.GELU(approximate="tanh"),
nn.Linear(4 * cfg["emb_dim"], cfg["emb_dim"]),
)
def forward(self, x):
return self.layers(x)
class TransformerBlockFast(nn.Module):
def __init__(self, cfg):
super().__init__()
self.att = PyTorchMultiHeadAttention(
d_in=cfg["emb_dim"],
d_out=cfg["emb_dim"],
num_heads=cfg["n_heads"],
dropout=cfg["drop_rate"],
qkv_bias=cfg["qkv_bias"])
self.ff = FeedForwardFast(cfg)
self.norm1 = nn.LayerNorm(cfg["emb_dim"])
self.norm2 = nn.LayerNorm(cfg["emb_dim"])
self.drop_shortcut = nn.Dropout(cfg["drop_rate"])
def forward(self, x):
# Shortcut connection for attention block
shortcut = x
x = self.norm1(x)
x = self.att(x) # Shape [batch_size, num_tokens, emb_size]
x = self.drop_shortcut(x)
x = x + shortcut # Add the original input back
# Shortcut connection for feed-forward block
shortcut = x
x = self.norm2(x)
x = self.ff(x)
x = self.drop_shortcut(x)
x = x + shortcut # Add the original input back
return x
class GPTModelFast(nn.Module):
"""
A faster variant of GPTModel optimized for training speed.
This version is only marginally faster on CPU (~1.02x) but significantly
faster on GPU (~2.05x) during training, thanks to optimized CUDA kernels
and FlashAttention support.
Key differences from the original GPTModel:
1. Uses PyTorch's built-in LayerNorm instead of a custom implementation.
2. Uses PyTorch's built-in GELU instead of a custom implementation.
3. Uses PyTorch's scaled_dot_product_attention instead of a custom MultiHeadAttention.
4. Automatically enables FlashAttention on compatible GPUs.
"""
def __init__(self, cfg):
super().__init__()
self.tok_emb = nn.Embedding(cfg["vocab_size"], cfg["emb_dim"])
self.pos_emb = nn.Embedding(cfg["context_length"], cfg["emb_dim"])
self.drop_emb = nn.Dropout(cfg["drop_rate"])
self.trf_blocks = nn.Sequential(
*[TransformerBlockFast(cfg) for _ in range(cfg["n_layers"])])
self.final_norm = nn.LayerNorm(cfg["emb_dim"])
self.out_head = nn.Linear(cfg["emb_dim"], cfg["vocab_size"], bias=False)
def forward(self, in_idx):
batch_size, seq_len = in_idx.shape
tok_embeds = self.tok_emb(in_idx)
pos_embeds = self.pos_emb(torch.arange(seq_len, device=in_idx.device))
x = tok_embeds + pos_embeds
x = self.drop_emb(x)
x = self.trf_blocks(x)
x = self.final_norm(x)
logits = self.out_head(x)
return logits