121 lines
5.0 KiB
Python
121 lines
5.0 KiB
Python
import inspect
|
|
from argparse import Namespace
|
|
|
|
import warnings
|
|
|
|
import torch
|
|
from torch import nn
|
|
|
|
from einops import rearrange, repeat
|
|
|
|
from ml_lib.metrics.multi_class_classification import MultiClassScores
|
|
from ml_lib.modules.blocks import TransformerModule
|
|
from ml_lib.modules.util import (LightningBaseModule, AutoPadToShape, F_x)
|
|
from util.module_mixins import CombinedModelMixins
|
|
|
|
MIN_NUM_PATCHES = 16
|
|
|
|
|
|
class VisualTransformer(CombinedModelMixins,
|
|
LightningBaseModule
|
|
):
|
|
|
|
def __init__(self, in_shape, n_classes, weight_init, activation,
|
|
embedding_size, heads, attn_depth, patch_size, use_residual,
|
|
use_bias, use_norm, dropout, lat_dim, loss, scheduler, mlp_dim, head_dim,
|
|
lr, weight_decay, sto_weight_avg, lr_scheduler_parameter, opt_reset_interval):
|
|
|
|
# TODO: Move this to parent class, or make it much easieer to access... But How...
|
|
a = dict(locals())
|
|
params = {arg: a[arg] for arg in inspect.signature(self.__init__).parameters.keys() if arg != 'self'}
|
|
super(VisualTransformer, self).__init__(params)
|
|
|
|
self.in_shape = in_shape
|
|
assert len(self.in_shape) == 3, 'There need to be three Dimensions'
|
|
channels, height, width = self.in_shape
|
|
|
|
# Model Paramters
|
|
# =============================================================================
|
|
# Additional parameters
|
|
self.embed_dim = self.params.embedding_size
|
|
|
|
# Automatic Image Shaping
|
|
self.patch_size = self.params.patch_size
|
|
image_size = (max(height, width) // self.patch_size) * self.patch_size
|
|
self.image_size = image_size + self.patch_size if image_size < max(height, width) else image_size
|
|
|
|
# This should be obsolete
|
|
assert self.image_size % self.patch_size == 0, 'image dimensions must be divisible by the patch size'
|
|
|
|
num_patches = (self.image_size // self.patch_size) ** 2
|
|
patch_dim = channels * self.patch_size ** 2
|
|
assert num_patches >= MIN_NUM_PATCHES, f'your number of patches ({num_patches}) is way too small for ' + \
|
|
f'attention. Try decreasing your patch size'
|
|
|
|
# Correct the Embedding Dim
|
|
#if not self.embed_dim % self.params.heads == 0:
|
|
# self.embed_dim = (self.embed_dim // self.params.heads) * self.params.heads
|
|
# message = ('Embedding Dimension was fixed to be devideable by the number' +
|
|
# f' of attention heads, is now: {self.embed_dim}')
|
|
# for func in print, warnings.warn:
|
|
# func(message)
|
|
|
|
# Utility Modules
|
|
self.autopad = AutoPadToShape((self.image_size, self.image_size))
|
|
|
|
# Modules with Parameters
|
|
self.transformer = TransformerModule(in_shape=self.embed_dim, mlp_dim=self.params.mlp_dim,
|
|
head_dim=self.params.head_dim,
|
|
heads=self.params.heads, depth=self.params.attn_depth,
|
|
dropout=self.params.dropout, use_norm=self.params.use_norm,
|
|
activation=self.params.activation, use_residual=self.params.use_residual
|
|
)
|
|
|
|
self.pos_embedding = nn.Parameter(torch.randn(1, num_patches + 1, self.embed_dim))
|
|
self.patch_to_embedding = nn.Linear(patch_dim, self.embed_dim)
|
|
self.cls_token = nn.Parameter(torch.randn(1, 1, self.embed_dim))
|
|
self.dropout = nn.Dropout(self.params.dropout)
|
|
|
|
self.to_cls_token = nn.Identity()
|
|
|
|
self.mlp_head = nn.Sequential(
|
|
nn.LayerNorm(self.embed_dim),
|
|
nn.Linear(self.embed_dim, self.params.lat_dim),
|
|
nn.GELU(),
|
|
nn.Dropout(self.params.dropout),
|
|
nn.Linear(self.params.lat_dim, n_classes),
|
|
nn.Softmax()
|
|
)
|
|
|
|
def forward(self, x, mask=None, return_attn_weights=False):
|
|
"""
|
|
:param x: the sequence to the encoder (required).
|
|
:param mask: the mask for the src sequence (optional).
|
|
:return:
|
|
"""
|
|
tensor = self.autopad(x)
|
|
p = self.params.patch_size
|
|
tensor = rearrange(tensor, 'b c (h p1) (w p2) -> b (h w) (p1 p2 c)', p1=p, p2=p)
|
|
|
|
tensor = self.patch_to_embedding(tensor)
|
|
b, n, _ = tensor.shape
|
|
|
|
cls_tokens = repeat(self.cls_token, '() n d -> b n d', b=b)
|
|
|
|
tensor = torch.cat((cls_tokens, tensor), dim=1)
|
|
tensor += self.pos_embedding[:, :(n + 1)]
|
|
tensor = self.dropout(tensor)
|
|
|
|
if return_attn_weights:
|
|
tensor, attn_weights = self.transformer(tensor, mask, return_attn_weights)
|
|
else:
|
|
attn_weights = None
|
|
tensor = self.transformer(tensor, mask)
|
|
|
|
tensor = self.to_cls_token(tensor[:, 0])
|
|
tensor = self.mlp_head(tensor)
|
|
return Namespace(main_out=tensor, attn_weights=attn_weights)
|
|
|
|
def additional_scores(self, outputs):
|
|
return MultiClassScores(self)(outputs)
|