Spaces:
Sleeping
Sleeping
File size: 2,340 Bytes
82fea12 |
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 |
# Copyright (c) Facebook, Inc. and its affiliates.
#
# This source code is licensed under the MIT license found in the
# LICENSE file in the root directory of this source tree.
from bitsandbytes.optim.optimizer import Optimizer1State
class SGD(Optimizer1State):
def __init__(
self,
params,
lr,
momentum=0,
dampening=0,
weight_decay=0,
nesterov=False,
optim_bits=32,
args=None,
min_8bit_size=4096,
percentile_clipping=100,
block_wise=True,
):
if momentum == 0:
raise NotImplementedError("SGD without momentum is not supported!")
super().__init__(
"momentum",
params,
lr,
(momentum, dampening),
0.0,
weight_decay,
optim_bits,
args,
min_8bit_size,
percentile_clipping,
block_wise,
)
class SGD8bit(Optimizer1State):
def __init__(
self,
params,
lr,
momentum=0,
dampening=0,
weight_decay=0,
nesterov=False,
args=None,
min_8bit_size=4096,
percentile_clipping=100,
block_wise=True,
):
if momentum == 0:
raise NotImplementedError("SGD without momentum is not supported!")
super().__init__(
"momentum",
params,
lr,
(momentum, dampening),
0.0,
weight_decay,
8,
args,
min_8bit_size,
percentile_clipping,
block_wise,
)
class SGD32bit(Optimizer1State):
def __init__(
self,
params,
lr,
momentum=0,
dampening=0,
weight_decay=0,
nesterov=False,
args=None,
min_8bit_size=4096,
percentile_clipping=100,
block_wise=True,
):
if momentum == 0:
raise NotImplementedError("SGD without momentum is not supported!")
super().__init__(
"momentum",
params,
lr,
(momentum, dampening),
0.0,
weight_decay,
32,
args,
min_8bit_size,
percentile_clipping,
block_wise,
)
|