Update quantization
Browse files- quantization.py +5 -1
quantization.py
CHANGED
|
@@ -7,7 +7,11 @@ import base64
|
|
| 7 |
import ctypes
|
| 8 |
|
| 9 |
from typing import List
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
|
| 12 |
|
| 13 |
class W8A16Linear(torch.autograd.Function):
|
|
|
|
| 7 |
import ctypes
|
| 8 |
|
| 9 |
from typing import List
|
| 10 |
+
|
| 11 |
+
try:
|
| 12 |
+
from cpm_kernels.kernels.base import LazyKernelCModule, KernelFunction, round_up
|
| 13 |
+
except:
|
| 14 |
+
pass
|
| 15 |
|
| 16 |
|
| 17 |
class W8A16Linear(torch.autograd.Function):
|