-
Notifications
You must be signed in to change notification settings - Fork 22
/
stip_llama.py
144 lines (115 loc) · 4.35 KB
/
stip_llama.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
"""
reference: https://github.com/facebookresearch/codellama
"""
import torch
import torch.nn as nn
import torch.nn.functional as F
import numpy as np
from stip_original import MultiHeadAttention
import copy
class FeedForward(nn.Module):
def __init__(
self,
dim: int,
hidden_dim: int,
):
super(FeedForward, self).__init__()
self.w1 = nn.Linear(dim, hidden_dim, bias=False)
self.w2 = nn.Linear(hidden_dim, dim, bias=False)
self.w3 = nn.Linear(dim, hidden_dim, bias=False)
def forward(self, x):
return self.w2(F.silu(self.w1(x)) * self.w3(x))
class RMSNorm(torch.nn.Module):
def __init__(self, dim: int, eps: float = 1e-6):
"""
Initialize the RMSNorm normalization layer.
Args:
dim (int): The dimension of the input tensor.
eps (float, optional): A small value added to the denominator for numerical stability. Default is 1e-6.
Attributes:
eps (float): A small value added to the denominator for numerical stability.
weight (nn.Parameter): Learnable scaling parameter.
"""
super().__init__()
self.eps = eps
self.weight = nn.Parameter(torch.ones(dim))
def _norm(self, x):
"""
Apply the RMSNorm normalization to the input tensor.
Args:
x (torch.Tensor): The input tensor.
Returns:
torch.Tensor: The normalized tensor.
"""
return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps)
def forward(self, x):
"""
Forward pass through the RMSNorm layer.
Args:
x (torch.Tensor): The input tensor.
Returns:
torch.Tensor: The output tensor after applying RMSNorm.
"""
output = self._norm(x.float()).type_as(x)
return output * self.weight
class LlamaTransformerBlock(nn.Module):
"""
"""
def __init__(self, d_model, num_heads, d_ff):
super(LlamaTransformerBlock, self).__init__()
self.attention_layer = MultiHeadAttention(d_model, num_heads)
self.ff_layer = FeedForward(d_model, d_ff)
self.rmsn1 = RMSNorm(d_model)
self.rmsn2 = RMSNorm(d_model)
def forward(self, x, mask):
attention_input = self.rmsn1(x)
attention_output = self.attention_layer(attention_input, attention_input, attention_input, mask)
h = x + attention_output
ff_input = self.rmsn2(h)
ff_output = self.ff_layer(ff_input)
res = h + ff_output
return res
def permute_block(blk, p):
p_blk = copy.deepcopy(blk)
with torch.no_grad():
for name, para in p_blk.named_parameters():
# print(name)
if name in ["attention_layer.w_q.weight",
"attention_layer.w_k.weight",
"attention_layer.w_v.weight",
"ff_layer.w1.weight",
"ff_layer.w3.weight"]:
para.data = para.data[:, p]
if name in ["attention_layer.w_o.weight",
"attention_layer.w_o.bias",
"ff_layer.w2.weight"]:
para.data = para.data[p]
return p_blk
if __name__ == "__main__":
BS = 2
SEQLEN = 3
DMODEL = 4
DFF = 8
NHEADS = 2
TEST_BLK = 1
if TEST_BLK:
x = torch.from_numpy(np.random.rand(BS, SEQLEN, DMODEL)).float()
block = LlamaTransformerBlock(DMODEL, NHEADS, DFF)
p = np.random.permutation(DMODEL)
print(f"Permutation={p}")
xp = x[:, :, p]
p_block = permute_block(block, p)
mask = (1 - torch.triu(torch.ones(1, SEQLEN, SEQLEN), diagonal=1)).bool()
with torch.no_grad():
print("Encoder Block:")
y = block(x, None)
yp = p_block(xp, None)
diff = np.abs(y[:, :, p] - yp).sum()
print("Original reslut:\n", y, "\nNew result:\n", yp)
print("Diff=", diff)
print("Decoder Block:")
y = block(x, mask)
yp = p_block(xp, mask)
diff = np.abs(y[:, :, p] - yp).sum()
print("Original reslut:\n", y, "\nNew result:\n", yp)
print("Diff=", diff)