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
|
# flake8: noqa: E266, C417, B950
from mixtral_moe_model import ConditionalFeedForward
import torch
import torch.nn as nn
import torch.nn.functional as F
##### Quantization Primitives ######
def dynamically_quantize_per_channel(x, quant_min, quant_max, target_dtype):
# assumes symmetric quantization
# assumes axis == 0
# assumes dense memory format
# TODO(future): relax ^ as needed
# default setup for affine quantization of activations
eps = torch.finfo(torch.float32).eps
# get min and max
min_val, max_val = torch.aminmax(x, dim=1)
# calculate scales and zero_points based on min and max
# reference: https://fburl.com/code/srbiybme
min_val_neg = torch.min(min_val, torch.zeros_like(min_val))
max_val_pos = torch.max(max_val, torch.zeros_like(max_val))
device = min_val_neg.device
# reference: https://fburl.com/code/4wll53rk
max_val_pos = torch.max(-min_val_neg, max_val_pos)
scales = max_val_pos / (float(quant_max - quant_min) / 2)
# ensure scales is the same dtype as the original tensor
scales = torch.clamp(scales, min=eps).to(x.dtype)
zero_points = torch.zeros(min_val_neg.size(), dtype=torch.int64, device=device)
# quantize based on qmin/qmax/scales/zp
# reference: https://www.internalfb.com/code/fbsource/[8edc275012b1]/fbcode/caffe2/torch/ao/quantization/fx/_decomposed.py?lines=63
x_div = x / scales.unsqueeze(-1)
x_round = torch.round(x_div)
x_zp = x_round + zero_points.unsqueeze(-1)
quant = torch.clamp(x_zp, quant_min, quant_max).to(target_dtype)
return quant, scales, zero_points
##### Weight-only int8 per-channel quantized code ######
def replace_linear_weight_only_int8_per_channel(module):
for name, child in module.named_children():
if isinstance(child, nn.Linear) and name != "gate":
setattr(
module,
name,
WeightOnlyInt8Linear(
child.in_features, child.out_features, target_dtype=torch.int8
),
)
elif isinstance(child, ConditionalFeedForward):
num_experts, intermediate_size, dim = child.w1.shape
setattr(
module,
name,
ConditionalFeedForwardInt8(
num_experts, intermediate_size, dim, target_dtype=torch.int8
),
)
else:
replace_linear_weight_only_int8_per_channel(child)
class WeightOnlyInt8QuantHandler:
def __init__(self, mod):
self.mod = mod
@torch.no_grad()
def create_quantized_state_dict(self):
cur_state_dict = self.mod.state_dict()
for fqn, mod in self.mod.named_modules():
if isinstance(mod, torch.nn.Linear) and not fqn.endswith(".gate"):
int8_weight, scales, _ = dynamically_quantize_per_channel(
mod.weight.float(), -128, 127, torch.int8
)
cur_state_dict[f"{fqn}.weight"] = int8_weight
cur_state_dict[f"{fqn}.scales"] = scales.to(mod.weight.dtype)
elif isinstance(mod, ConditionalFeedForward):
for weight_idx in range(0, 3):
weight_name = f"w{weight_idx + 1}"
scales_name = f"scales{weight_idx + 1}"
weight = getattr(mod, weight_name)
num_experts, intermediate_size, dim = weight.shape
bit8_weight_list = []
scales_list = []
for expert_idx in range(num_experts):
bit8_weight, scales, _ = dynamically_quantize_per_channel(
weight[expert_idx].float(), -128, 127, torch.int8
)
bit8_weight_list.append(
bit8_weight.reshape(1, intermediate_size, dim)
)
scales_list.append(scales.reshape(1, intermediate_size))
cur_state_dict[f"{fqn}.{weight_name}"] = torch.cat(
bit8_weight_list, dim=0
)
cur_state_dict[f"{fqn}.{scales_name}"] = torch.cat(
scales_list, dim=0
)
return cur_state_dict
def convert_for_runtime(self):
replace_linear_weight_only_int8_per_channel(self.mod)
return self.mod
class WeightOnlyInt8Linear(torch.nn.Module):
__constants__ = ["in_features", "out_features"]
in_features: int
out_features: int
weight: torch.Tensor
def __init__(
self,
in_features: int,
out_features: int,
bias: bool = True,
device=None,
dtype=None,
target_dtype=None,
) -> None:
assert target_dtype is not None
super().__init__()
self.in_features = in_features
self.out_features = out_features
self.register_buffer(
"weight", torch.empty((out_features, in_features), dtype=target_dtype)
)
self.register_buffer("scales", torch.ones(out_features, dtype=torch.bfloat16))
def forward(self, input: torch.Tensor) -> torch.Tensor:
return F.linear(input, self.weight.to(dtype=input.dtype)) * self.scales
class ConditionalFeedForwardInt8(nn.Module):
def __init__(self, num_experts, intermediate_size, dim, target_dtype):
super().__init__()
self.target_dtype = target_dtype
self.register_buffer(
"w1", torch.empty(num_experts, intermediate_size, dim, dtype=target_dtype)
)
self.register_buffer(
"w2", torch.empty(num_experts, dim, intermediate_size, dtype=target_dtype)
)
self.register_buffer(
"w3", torch.empty(num_experts, intermediate_size, dim, dtype=target_dtype)
)
self.register_buffer(
"scales1", torch.empty(num_experts, intermediate_size, dtype=torch.bfloat16)
)
self.register_buffer(
"scales2", torch.empty(num_experts, dim, dtype=torch.bfloat16)
)
self.register_buffer(
"scales3", torch.empty(num_experts, intermediate_size, dtype=torch.bfloat16)
)
def forward(self, x, expert_indices):
w1_weights = self.w1.to(x.dtype)[expert_indices] # [T, A, D, D]
w3_weights = self.w3.to(x.dtype)[expert_indices] # [T, A, D, D]
w2_weights = self.w2.to(x.dtype)[expert_indices]
x1 = F.silu(
torch.einsum("ti,taoi -> tao", x, w1_weights)
* self.scales1[expert_indices].to(x.dtype)
)
x3 = torch.einsum("ti, taoi -> tao", x, w3_weights) * self.scales3[
expert_indices
].to(x.dtype)
expert_outs = torch.einsum(
"tao, taio -> tai", (x1 * x3), w2_weights
) * self.scales2[expert_indices].to(x.dtype) # [T, A, D, D]
return expert_outs
|
