226 lines
8.6 KiB
Python
226 lines
8.6 KiB
Python
""" DropBlock, DropPath
|
|
|
|
PyTorch implementations of DropBlock and DropPath (Stochastic Depth) regularization layers.
|
|
|
|
Papers:
|
|
DropBlock: A regularization method for convolutional networks (https://arxiv.org/abs/1810.12890)
|
|
|
|
Deep Networks with Stochastic Depth (https://arxiv.org/abs/1603.09382)
|
|
|
|
Code:
|
|
DropBlock impl inspired by two Tensorflow impl that I liked:
|
|
- https://github.com/tensorflow/tpu/blob/master/models/official/resnet/resnet_model.py#L74
|
|
- https://github.com/clovaai/assembled-cnn/blob/master/nets/blocks.py
|
|
|
|
Hacked together by / Copyright 2020 Ross Wightman
|
|
"""
|
|
from typing import List, Union
|
|
|
|
import torch
|
|
import torch.nn as nn
|
|
import torch.nn.functional as F
|
|
|
|
from .grid import ndgrid
|
|
|
|
|
|
def drop_block_2d(
|
|
x,
|
|
drop_prob: float = 0.1,
|
|
block_size: int = 7,
|
|
gamma_scale: float = 1.0,
|
|
with_noise: bool = False,
|
|
inplace: bool = False,
|
|
batchwise: bool = False
|
|
):
|
|
""" DropBlock. See https://arxiv.org/pdf/1810.12890.pdf
|
|
|
|
DropBlock with an experimental gaussian noise option. This layer has been tested on a few training
|
|
runs with success, but needs further validation and possibly optimization for lower runtime impact.
|
|
"""
|
|
B, C, H, W = x.shape
|
|
total_size = W * H
|
|
clipped_block_size = min(block_size, min(W, H))
|
|
# seed_drop_rate, the gamma parameter
|
|
gamma = gamma_scale * drop_prob * total_size / clipped_block_size ** 2 / (
|
|
(W - block_size + 1) * (H - block_size + 1))
|
|
|
|
# Forces the block to be inside the feature map.
|
|
w_i, h_i = ndgrid(torch.arange(W, device=x.device), torch.arange(H, device=x.device))
|
|
valid_block = ((w_i >= clipped_block_size // 2) & (w_i < W - (clipped_block_size - 1) // 2)) & \
|
|
((h_i >= clipped_block_size // 2) & (h_i < H - (clipped_block_size - 1) // 2))
|
|
valid_block = torch.reshape(valid_block, (1, 1, H, W)).to(dtype=x.dtype)
|
|
|
|
if batchwise:
|
|
# one mask for whole batch, quite a bit faster
|
|
uniform_noise = torch.rand((1, C, H, W), dtype=x.dtype, device=x.device)
|
|
else:
|
|
uniform_noise = torch.rand_like(x)
|
|
block_mask = ((2 - gamma - valid_block + uniform_noise) >= 1).to(dtype=x.dtype)
|
|
block_mask = -F.max_pool2d(
|
|
-block_mask,
|
|
kernel_size=clipped_block_size, # block_size,
|
|
stride=1,
|
|
padding=clipped_block_size // 2)
|
|
|
|
if with_noise:
|
|
normal_noise = torch.randn((1, C, H, W), dtype=x.dtype, device=x.device) if batchwise else torch.randn_like(x)
|
|
if inplace:
|
|
x.mul_(block_mask).add_(normal_noise * (1 - block_mask))
|
|
else:
|
|
x = x * block_mask + normal_noise * (1 - block_mask)
|
|
else:
|
|
normalize_scale = (block_mask.numel() / block_mask.to(dtype=torch.float32).sum().add(1e-7)).to(x.dtype)
|
|
if inplace:
|
|
x.mul_(block_mask * normalize_scale)
|
|
else:
|
|
x = x * block_mask * normalize_scale
|
|
return x
|
|
|
|
|
|
def drop_block_fast_2d(
|
|
x: torch.Tensor,
|
|
drop_prob: float = 0.1,
|
|
block_size: int = 7,
|
|
gamma_scale: float = 1.0,
|
|
with_noise: bool = False,
|
|
inplace: bool = False,
|
|
):
|
|
""" DropBlock. See https://arxiv.org/pdf/1810.12890.pdf
|
|
|
|
DropBlock with an experimental gaussian noise option. Simplied from above without concern for valid
|
|
block mask at edges.
|
|
"""
|
|
B, C, H, W = x.shape
|
|
total_size = W * H
|
|
clipped_block_size = min(block_size, min(W, H))
|
|
gamma = gamma_scale * drop_prob * total_size / clipped_block_size ** 2 / (
|
|
(W - block_size + 1) * (H - block_size + 1))
|
|
|
|
block_mask = torch.empty_like(x).bernoulli_(gamma)
|
|
block_mask = F.max_pool2d(
|
|
block_mask.to(x.dtype), kernel_size=clipped_block_size, stride=1, padding=clipped_block_size // 2)
|
|
|
|
if with_noise:
|
|
normal_noise = torch.empty_like(x).normal_()
|
|
if inplace:
|
|
x.mul_(1. - block_mask).add_(normal_noise * block_mask)
|
|
else:
|
|
x = x * (1. - block_mask) + normal_noise * block_mask
|
|
else:
|
|
block_mask = 1 - block_mask
|
|
normalize_scale = (block_mask.numel() / block_mask.to(dtype=torch.float32).sum().add(1e-6)).to(dtype=x.dtype)
|
|
if inplace:
|
|
x.mul_(block_mask * normalize_scale)
|
|
else:
|
|
x = x * block_mask * normalize_scale
|
|
return x
|
|
|
|
|
|
class DropBlock2d(nn.Module):
|
|
""" DropBlock. See https://arxiv.org/pdf/1810.12890.pdf
|
|
"""
|
|
|
|
def __init__(
|
|
self,
|
|
drop_prob: float = 0.1,
|
|
block_size: int = 7,
|
|
gamma_scale: float = 1.0,
|
|
with_noise: bool = False,
|
|
inplace: bool = False,
|
|
batchwise: bool = False,
|
|
fast: bool = True):
|
|
super().__init__()
|
|
self.drop_prob = drop_prob
|
|
self.gamma_scale = gamma_scale
|
|
self.block_size = block_size
|
|
self.with_noise = with_noise
|
|
self.inplace = inplace
|
|
self.batchwise = batchwise
|
|
self.fast = fast # FIXME finish comparisons of fast vs not
|
|
|
|
def forward(self, x):
|
|
if not self.training or not self.drop_prob:
|
|
return x
|
|
if self.fast:
|
|
return drop_block_fast_2d(
|
|
x, self.drop_prob, self.block_size, self.gamma_scale, self.with_noise, self.inplace)
|
|
else:
|
|
return drop_block_2d(
|
|
x, self.drop_prob, self.block_size, self.gamma_scale, self.with_noise, self.inplace, self.batchwise)
|
|
|
|
|
|
def drop_path(x, drop_prob: float = 0., training: bool = False, scale_by_keep: bool = True):
|
|
"""Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
|
|
|
|
This is the same as the DropConnect impl I created for EfficientNet, etc networks, however,
|
|
the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper...
|
|
See discussion: https://github.com/tensorflow/tpu/issues/494#issuecomment-532968956 ... I've opted for
|
|
changing the layer and argument names to 'drop path' rather than mix DropConnect as a layer name and use
|
|
'survival rate' as the argument.
|
|
|
|
"""
|
|
if drop_prob == 0. or not training:
|
|
return x
|
|
keep_prob = 1 - drop_prob
|
|
shape = (x.shape[0],) + (1,) * (x.ndim - 1) # work with diff dim tensors, not just 2D ConvNets
|
|
random_tensor = x.new_empty(shape).bernoulli_(keep_prob)
|
|
if keep_prob > 0.0 and scale_by_keep:
|
|
random_tensor.div_(keep_prob)
|
|
return x * random_tensor
|
|
|
|
|
|
class DropPath(nn.Module):
|
|
"""Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
|
|
"""
|
|
def __init__(self, drop_prob: float = 0., scale_by_keep: bool = True):
|
|
super().__init__()
|
|
self.drop_prob = drop_prob
|
|
self.scale_by_keep = scale_by_keep
|
|
|
|
def forward(self, x):
|
|
return drop_path(x, self.drop_prob, self.training, self.scale_by_keep)
|
|
|
|
def extra_repr(self):
|
|
return f'drop_prob={round(self.drop_prob,3):0.3f}'
|
|
|
|
|
|
def calculate_drop_path_rates(
|
|
drop_path_rate: float,
|
|
depths: Union[int, List[int]],
|
|
stagewise: bool = False,
|
|
) -> Union[List[float], List[List[float]]]:
|
|
"""Generate drop path rates for stochastic depth.
|
|
|
|
This function handles two common patterns for drop path rate scheduling:
|
|
1. Per-block: Linear increase from 0 to drop_path_rate across all blocks
|
|
2. Stage-wise: Linear increase across stages, with same rate within each stage
|
|
|
|
Args:
|
|
drop_path_rate: Maximum drop path rate (at the end).
|
|
depths: Either a single int for total depth (per-block mode) or
|
|
list of ints for depths per stage (stage-wise mode).
|
|
stagewise: If True, use stage-wise pattern. If False, use per-block pattern.
|
|
When depths is a list, stagewise defaults to True.
|
|
|
|
Returns:
|
|
For per-block mode: List of drop rates, one per block.
|
|
For stage-wise mode: List of lists, drop rates per stage.
|
|
"""
|
|
if isinstance(depths, int):
|
|
# Single depth value - per-block pattern
|
|
if stagewise:
|
|
raise ValueError("stagewise=True requires depths to be a list of stage depths")
|
|
dpr = [x.item() for x in torch.linspace(0, drop_path_rate, depths, device='cpu')]
|
|
return dpr
|
|
else:
|
|
# List of depths - can be either pattern
|
|
total_depth = sum(depths)
|
|
if stagewise:
|
|
# Stage-wise pattern: same drop rate within each stage
|
|
dpr = [x.tolist() for x in torch.linspace(0, drop_path_rate, total_depth, device='cpu').split(depths)]
|
|
return dpr
|
|
else:
|
|
# Per-block pattern across all stages
|
|
dpr = [x.item() for x in torch.linspace(0, drop_path_rate, total_depth, device='cpu')]
|
|
return dpr
|