Fully move ViT hybrids to their own file, including embedding module. Remove some extra DeiT models that were for benchmarking only.

pull/533/head
Ross Wightman 4 years ago
parent a5310a3451
commit ea9c9550b2

@ -5,6 +5,9 @@ A PyTorch implement of Vision Transformers as described in
The official jax code is released and available at https://github.com/google-research/vision_transformer The official jax code is released and available at https://github.com/google-research/vision_transformer
DeiT model defs and weights from https://github.com/facebookresearch/deit,
paper `DeiT: Data-efficient Image Transformers` - https://arxiv.org/abs/2012.12877
Acknowledgments: Acknowledgments:
* The paper authors for releasing code and weights, thanks! * The paper authors for releasing code and weights, thanks!
* I fixed my class token impl based on Phil Wang's https://github.com/lucidrains/vit-pytorch ... check it out * I fixed my class token impl based on Phil Wang's https://github.com/lucidrains/vit-pytorch ... check it out
@ -12,9 +15,6 @@ for some einops/einsum fun
* Simple transformer style inspired by Andrej Karpathy's https://github.com/karpathy/minGPT * Simple transformer style inspired by Andrej Karpathy's https://github.com/karpathy/minGPT
* Bert reference code checks against Huggingface Transformers and Tensorflow Bert * Bert reference code checks against Huggingface Transformers and Tensorflow Bert
DeiT model defs and weights from https://github.com/facebookresearch/deit,
paper `DeiT: Data-efficient Image Transformers` - https://arxiv.org/abs/2012.12877
Hacked together by / Copyright 2020 Ross Wightman Hacked together by / Copyright 2020 Ross Wightman
""" """
import math import math
@ -99,18 +99,8 @@ default_cfgs = {
# deit models (FB weights) # deit models (FB weights)
'vit_deit_tiny_patch16_224': _cfg( 'vit_deit_tiny_patch16_224': _cfg(
url='https://dl.fbaipublicfiles.com/deit/deit_tiny_patch16_224-a1311bcf.pth'), url='https://dl.fbaipublicfiles.com/deit/deit_tiny_patch16_224-a1311bcf.pth'),
'vit_deit_tiny_patch16_224_in21k': _cfg(num_classes=21843),
'vit_deit_tiny_patch16_384': _cfg(input_size=(3, 384, 384)),
'vit_deit_small_patch16_224': _cfg( 'vit_deit_small_patch16_224': _cfg(
url='https://dl.fbaipublicfiles.com/deit/deit_small_patch16_224-cd65a155.pth'), url='https://dl.fbaipublicfiles.com/deit/deit_small_patch16_224-cd65a155.pth'),
'vit_deit_small_patch16_224_in21k': _cfg(num_classes=21843),
'vit_deit_small_patch16_384': _cfg(input_size=(3, 384, 384)),
'vit_deit_small_patch32_224': _cfg(),
'vit_deit_small_patch32_224_in21k': _cfg(num_classes=21843),
'vit_deit_small_patch32_384': _cfg(input_size=(3, 384, 384)),
'vit_deit_base_patch16_224': _cfg( 'vit_deit_base_patch16_224': _cfg(
url='https://dl.fbaipublicfiles.com/deit/deit_base_patch16_224-b5f2ef4d.pth',), url='https://dl.fbaipublicfiles.com/deit/deit_base_patch16_224-b5f2ef4d.pth',),
'vit_deit_base_patch16_384': _cfg( 'vit_deit_base_patch16_384': _cfg(
@ -220,48 +210,6 @@ class PatchEmbed(nn.Module):
return x return x
class HybridEmbed(nn.Module):
""" CNN Feature Map Embedding
Extract feature map from CNN, flatten, project to embedding dim.
"""
def __init__(self, backbone, img_size=224, patch_size=1, feature_size=None, in_chans=3, embed_dim=768):
super().__init__()
assert isinstance(backbone, nn.Module)
img_size = to_2tuple(img_size)
patch_size = to_2tuple(patch_size)
self.img_size = img_size
self.patch_size = patch_size
self.backbone = backbone
if feature_size is None:
with torch.no_grad():
# NOTE Most reliable way of determining output dims is to run forward pass
training = backbone.training
if training:
backbone.eval()
o = self.backbone(torch.zeros(1, in_chans, img_size[0], img_size[1]))
if isinstance(o, (list, tuple)):
o = o[-1] # last feature if backbone outputs list/tuple of features
feature_size = o.shape[-2:]
feature_dim = o.shape[1]
backbone.train(training)
else:
feature_size = to_2tuple(feature_size)
if hasattr(self.backbone, 'feature_info'):
feature_dim = self.backbone.feature_info.channels()[-1]
else:
feature_dim = self.backbone.num_features
assert feature_size[0] % patch_size[0] == 0 and feature_size[1] % patch_size[1] == 0
self.num_patches = feature_size[0] // patch_size[0] * feature_size[1] // patch_size[1]
self.proj = nn.Conv2d(feature_dim, embed_dim, kernel_size=patch_size, stride=patch_size)
def forward(self, x):
x = self.backbone(x)
if isinstance(x, (list, tuple)):
x = x[-1] # last feature if backbone outputs list/tuple of features
x = self.proj(x).flatten(2).transpose(1, 2)
return x
class VisionTransformer(nn.Module): class VisionTransformer(nn.Module):
""" Vision Transformer """ Vision Transformer
@ -274,7 +222,7 @@ class VisionTransformer(nn.Module):
def __init__(self, img_size=224, patch_size=16, in_chans=3, num_classes=1000, embed_dim=768, depth=12, def __init__(self, img_size=224, patch_size=16, in_chans=3, num_classes=1000, embed_dim=768, depth=12,
num_heads=12, mlp_ratio=4., qkv_bias=True, qk_scale=None, representation_size=None, distilled=False, num_heads=12, mlp_ratio=4., qkv_bias=True, qk_scale=None, representation_size=None, distilled=False,
drop_rate=0., attn_drop_rate=0., drop_path_rate=0., hybrid_backbone=None, norm_layer=None, drop_rate=0., attn_drop_rate=0., drop_path_rate=0., embed_layer=PatchEmbed, norm_layer=None,
act_layer=None, weight_init=''): act_layer=None, weight_init=''):
""" """
Args: Args:
@ -293,7 +241,7 @@ class VisionTransformer(nn.Module):
drop_rate (float): dropout rate drop_rate (float): dropout rate
attn_drop_rate (float): attention dropout rate attn_drop_rate (float): attention dropout rate
drop_path_rate (float): stochastic depth rate drop_path_rate (float): stochastic depth rate
hybrid_backbone (nn.Module): CNN backbone to use in-place of PatchEmbed module embed_layer (nn.Module): patch embedding layer
norm_layer: (nn.Module): normalization layer norm_layer: (nn.Module): normalization layer
weight_init: (str): weight init scheme weight_init: (str): weight init scheme
""" """
@ -303,13 +251,8 @@ class VisionTransformer(nn.Module):
self.num_tokens = 2 if distilled else 1 self.num_tokens = 2 if distilled else 1
norm_layer = norm_layer or partial(nn.LayerNorm, eps=1e-6) norm_layer = norm_layer or partial(nn.LayerNorm, eps=1e-6)
act_layer = act_layer or nn.GELU act_layer = act_layer or nn.GELU
patch_size = patch_size or (1 if hybrid_backbone is not None else 16)
if hybrid_backbone is not None: self.patch_embed = embed_layer(
self.patch_embed = HybridEmbed(
hybrid_backbone, img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim)
else:
self.patch_embed = PatchEmbed(
img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim) img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim)
num_patches = self.patch_embed.num_patches num_patches = self.patch_embed.num_patches
@ -489,7 +432,8 @@ def checkpoint_filter_fn(state_dict, model):
return out_dict return out_dict
def _create_vision_transformer(variant, pretrained=False, **kwargs): def _create_vision_transformer(variant, pretrained=False, default_cfg=None, **kwargs):
if default_cfg is None:
default_cfg = deepcopy(default_cfgs[variant]) default_cfg = deepcopy(default_cfgs[variant])
overlay_external_default_cfg(default_cfg, kwargs) overlay_external_default_cfg(default_cfg, kwargs)
default_num_classes = default_cfg['num_classes'] default_num_classes = default_cfg['num_classes']
@ -680,22 +624,6 @@ def vit_deit_tiny_patch16_224(pretrained=False, **kwargs):
return model return model
@register_model
def vit_deit_tiny_patch16_224_in21k(pretrained=False, **kwargs):
""" DeiT-tiny model"""
model_kwargs = dict(patch_size=16, embed_dim=192, depth=12, num_heads=3, representation_size=192, **kwargs)
model = _create_vision_transformer('vit_deit_tiny_patch16_224_in21k', pretrained=pretrained, **model_kwargs)
return model
@register_model
def vit_deit_tiny_patch16_384(pretrained=False, **kwargs):
""" DeiT-tiny model"""
model_kwargs = dict(patch_size=16, embed_dim=192, depth=12, num_heads=3, **kwargs)
model = _create_vision_transformer('vit_deit_tiny_patch16_384', pretrained=pretrained, **model_kwargs)
return model
@register_model @register_model
def vit_deit_small_patch16_224(pretrained=False, **kwargs): def vit_deit_small_patch16_224(pretrained=False, **kwargs):
""" DeiT-small model @ 224x224 from paper (https://arxiv.org/abs/2012.12877). """ DeiT-small model @ 224x224 from paper (https://arxiv.org/abs/2012.12877).
@ -706,48 +634,6 @@ def vit_deit_small_patch16_224(pretrained=False, **kwargs):
return model return model
@register_model
def vit_deit_small_patch16_224_in21k(pretrained=False, **kwargs):
""" DeiT-small """
model_kwargs = dict(patch_size=16, embed_dim=384, depth=12, num_heads=6, representation_size=384, **kwargs)
model = _create_vision_transformer('vit_deit_small_patch16_224_in21k', pretrained=pretrained, **model_kwargs)
return model
@register_model
def vit_deit_small_patch16_384(pretrained=False, **kwargs):
""" DeiT-small """
model_kwargs = dict(patch_size=16, embed_dim=384, depth=12, num_heads=6, **kwargs)
model = _create_vision_transformer('vit_deit_small_patch16_384', pretrained=pretrained, **model_kwargs)
return model
@register_model
def vit_deit_small_patch32_224(pretrained=False, **kwargs):
""" DeiT-small model @ 224x224 from paper (https://arxiv.org/abs/2012.12877).
ImageNet-1k weights from https://github.com/facebookresearch/deit.
"""
model_kwargs = dict(patch_size=32, embed_dim=384, depth=12, num_heads=6, **kwargs)
model = _create_vision_transformer('vit_deit_small_patch32_224', pretrained=pretrained, **model_kwargs)
return model
@register_model
def vit_deit_small_patch32_224_in21k(pretrained=False, **kwargs):
""" DeiT-small """
model_kwargs = dict(patch_size=32, embed_dim=384, depth=12, num_heads=6, representation_size=384, **kwargs)
model = _create_vision_transformer('vit_deit_small_patch32_224_in21k', pretrained=pretrained, **model_kwargs)
return model
@register_model
def vit_deit_small_patch32_384(pretrained=False, **kwargs):
""" DeiT-small """
model_kwargs = dict(patch_size=32, embed_dim=384, depth=12, num_heads=6, **kwargs)
model = _create_vision_transformer('vit_deit_small_patch32_384', pretrained=pretrained, **model_kwargs)
return model
@register_model @register_model
def vit_deit_base_patch16_224(pretrained=False, **kwargs): def vit_deit_base_patch16_224(pretrained=False, **kwargs):
""" DeiT base model @ 224x224 from paper (https://arxiv.org/abs/2012.12877). """ DeiT base model @ 224x224 from paper (https://arxiv.org/abs/2012.12877).

@ -9,6 +9,12 @@ keep file sizes sane.
Hacked together by / Copyright 2020 Ross Wightman Hacked together by / Copyright 2020 Ross Wightman
""" """
from copy import deepcopy
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from .layers import StdConv2dSame, StdConv2d, to_2tuple from .layers import StdConv2dSame, StdConv2d, to_2tuple
from .resnet import resnet26d, resnet50d from .resnet import resnet26d, resnet50d
@ -41,39 +47,14 @@ default_cfgs = {
# hybrid in-1k models (mostly untrained, experimental configs w/ resnetv2 stdconv backbones) # hybrid in-1k models (mostly untrained, experimental configs w/ resnetv2 stdconv backbones)
'vit_tiny_r_s16_p8_224': _cfg(), 'vit_tiny_r_s16_p8_224': _cfg(),
'vit_tiny_r_s16_p8_384': _cfg( 'vit_small_r_s16_p8_224': _cfg(),
input_size=(3, 384, 384), crop_pct=1.0),
'vit_small_r_s16_p8_224': _cfg(
crop_pct=1.0),
'vit_small_r_s16_p8_384': _cfg(
input_size=(3, 384, 384), crop_pct=1.0),
'vit_small_r20_s16_p2_224': _cfg(), 'vit_small_r20_s16_p2_224': _cfg(),
'vit_small_r20_s16_p2_384': _cfg(
input_size=(3, 384, 384), crop_pct=1.0),
'vit_small_r20_s16_224': _cfg(), 'vit_small_r20_s16_224': _cfg(),
'vit_small_r20_s16_384': _cfg(
input_size=(3, 384, 384), crop_pct=1.0),
'vit_small_r26_s32_224': _cfg(), 'vit_small_r26_s32_224': _cfg(),
'vit_small_r26_s32_384': _cfg(
input_size=(3, 384, 384), crop_pct=1.0),
'vit_base_r20_s16_224': _cfg(), 'vit_base_r20_s16_224': _cfg(),
'vit_base_r20_s16_384': _cfg(
input_size=(3, 384, 384), crop_pct=1.0),
'vit_base_r26_s32_224': _cfg(), 'vit_base_r26_s32_224': _cfg(),
'vit_base_r26_s32_384': _cfg(
input_size=(3, 384, 384), crop_pct=1.0),
'vit_base_r50_s16_224': _cfg(), 'vit_base_r50_s16_224': _cfg(),
'vit_large_r50_s32_224': _cfg(), 'vit_large_r50_s32_224': _cfg(),
'vit_large_r50_s32_384': _cfg(
input_size=(3, 384, 384), crop_pct=1.0),
# hybrid models (using timm resnet backbones) # hybrid models (using timm resnet backbones)
'vit_small_resnet26d_224': _cfg(mean=IMAGENET_DEFAULT_MEAN, std=IMAGENET_DEFAULT_STD), 'vit_small_resnet26d_224': _cfg(mean=IMAGENET_DEFAULT_MEAN, std=IMAGENET_DEFAULT_STD),
@ -83,6 +64,56 @@ default_cfgs = {
} }
class HybridEmbed(nn.Module):
""" CNN Feature Map Embedding
Extract feature map from CNN, flatten, project to embedding dim.
"""
def __init__(self, backbone, img_size=224, patch_size=1, feature_size=None, in_chans=3, embed_dim=768):
super().__init__()
assert isinstance(backbone, nn.Module)
img_size = to_2tuple(img_size)
patch_size = to_2tuple(patch_size)
self.img_size = img_size
self.patch_size = patch_size
self.backbone = backbone
if feature_size is None:
with torch.no_grad():
# NOTE Most reliable way of determining output dims is to run forward pass
training = backbone.training
if training:
backbone.eval()
o = self.backbone(torch.zeros(1, in_chans, img_size[0], img_size[1]))
if isinstance(o, (list, tuple)):
o = o[-1] # last feature if backbone outputs list/tuple of features
feature_size = o.shape[-2:]
feature_dim = o.shape[1]
backbone.train(training)
else:
feature_size = to_2tuple(feature_size)
if hasattr(self.backbone, 'feature_info'):
feature_dim = self.backbone.feature_info.channels()[-1]
else:
feature_dim = self.backbone.num_features
assert feature_size[0] % patch_size[0] == 0 and feature_size[1] % patch_size[1] == 0
self.num_patches = feature_size[0] // patch_size[0] * feature_size[1] // patch_size[1]
self.proj = nn.Conv2d(feature_dim, embed_dim, kernel_size=patch_size, stride=patch_size)
def forward(self, x):
x = self.backbone(x)
if isinstance(x, (list, tuple)):
x = x[-1] # last feature if backbone outputs list/tuple of features
x = self.proj(x).flatten(2).transpose(1, 2)
return x
def _create_vision_transformer_hybrid(variant, backbone, pretrained=False, **kwargs):
default_cfg = deepcopy(default_cfgs[variant])
embed_layer = partial(HybridEmbed, backbone=backbone)
kwargs.setdefault('patch_size', 1) # default patch size for hybrid models if not set
return _create_vision_transformer(
variant, pretrained=pretrained, default_cfg=default_cfg, embed_layer=embed_layer, **kwargs)
def _resnetv2(layers=(3, 4, 9), **kwargs): def _resnetv2(layers=(3, 4, 9), **kwargs):
""" ResNet-V2 backbone helper""" """ ResNet-V2 backbone helper"""
padding_same = kwargs.get('padding_same', True) padding_same = kwargs.get('padding_same', True)
@ -108,9 +139,9 @@ def vit_base_r50_s16_224_in21k(pretrained=False, **kwargs):
ImageNet-21k weights @ 224x224, source https://github.com/google-research/vision_transformer. ImageNet-21k weights @ 224x224, source https://github.com/google-research/vision_transformer.
""" """
backbone = _resnetv2(layers=(3, 4, 9), **kwargs) backbone = _resnetv2(layers=(3, 4, 9), **kwargs)
model_kwargs = dict( model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, representation_size=768, **kwargs)
embed_dim=768, depth=12, num_heads=12, hybrid_backbone=backbone, representation_size=768, **kwargs) model = _create_vision_transformer_hybrid(
model = _create_vision_transformer('vit_base_r50_s16_224_in21k', pretrained=pretrained, **model_kwargs) 'vit_base_r50_s16_224_in21k', backbone=backbone, pretrained=pretrained, **model_kwargs)
return model return model
@ -120,8 +151,9 @@ def vit_base_r50_s16_384(pretrained=False, **kwargs):
ImageNet-1k weights fine-tuned from in21k @ 384x384, source https://github.com/google-research/vision_transformer. ImageNet-1k weights fine-tuned from in21k @ 384x384, source https://github.com/google-research/vision_transformer.
""" """
backbone = _resnetv2((3, 4, 9), **kwargs) backbone = _resnetv2((3, 4, 9), **kwargs)
model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, hybrid_backbone=backbone, **kwargs) model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, **kwargs)
model = _create_vision_transformer('vit_base_r50_s16_384', pretrained=pretrained, **model_kwargs) model = _create_vision_transformer_hybrid(
'vit_base_r50_s16_384', backbone=backbone, pretrained=pretrained, **model_kwargs)
return model return model
@ -130,20 +162,9 @@ def vit_tiny_r_s16_p8_224(pretrained=False, **kwargs):
""" R+ViT-Ti/S16 w/ 8x8 patch hybrid @ 224 x 224. """ R+ViT-Ti/S16 w/ 8x8 patch hybrid @ 224 x 224.
""" """
backbone = _resnetv2(layers=(), **kwargs) backbone = _resnetv2(layers=(), **kwargs)
model_kwargs = dict( model_kwargs = dict(patch_size=8, embed_dim=192, depth=12, num_heads=3, **kwargs)
patch_size=8, embed_dim=192, depth=12, num_heads=3, hybrid_backbone=backbone, **kwargs) model = _create_vision_transformer_hybrid(
model = _create_vision_transformer('vit_tiny_r_s16_p8_224', pretrained=pretrained, **model_kwargs) 'vit_tiny_r_s16_p8_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
return model
@register_model
def vit_tiny_r_s16_p8_384(pretrained=False, **kwargs):
""" R+ViT-Ti/S16 w/ 8x8 patch hybrid @ 224 x 224.
"""
backbone = _resnetv2(layers=(), **kwargs)
model_kwargs = dict(
patch_size=8, embed_dim=192, depth=12, num_heads=3, hybrid_backbone=backbone, **kwargs)
model = _create_vision_transformer('vit_tiny_r_s16_p8_384', pretrained=pretrained, **model_kwargs)
return model return model
@ -152,43 +173,21 @@ def vit_small_r_s16_p8_224(pretrained=False, **kwargs):
""" R+ViT-S/S16 w/ 8x8 patch hybrid @ 224 x 224. """ R+ViT-S/S16 w/ 8x8 patch hybrid @ 224 x 224.
""" """
backbone = _resnetv2(layers=(), **kwargs) backbone = _resnetv2(layers=(), **kwargs)
model_kwargs = dict( model_kwargs = dict(patch_size=8, embed_dim=384, depth=12, num_heads=6, **kwargs)
patch_size=8, embed_dim=384, depth=12, num_heads=6, hybrid_backbone=backbone, **kwargs) model = _create_vision_transformer_hybrid(
model = _create_vision_transformer('vit_small_r_s16_p8_224', pretrained=pretrained, **model_kwargs) 'vit_small_r_s16_p8_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
return model return model
@register_model
def vit_small_r_s16_p8_384(pretrained=False, **kwargs):
""" R+ViT-S/S16 w/ 8x8 patch hybrid @ 224 x 224.
"""
backbone = _resnetv2(layers=(), **kwargs)
model_kwargs = dict(
patch_size=8, embed_dim=384, depth=12, num_heads=6, hybrid_backbone=backbone, **kwargs)
model = _create_vision_transformer('vit_small_r_s16_p8_384', pretrained=pretrained, **model_kwargs)
return model
@register_model @register_model
def vit_small_r20_s16_p2_224(pretrained=False, **kwargs): def vit_small_r20_s16_p2_224(pretrained=False, **kwargs):
""" R52+ViT-S/S16 w/ 2x2 patch hybrid @ 224 x 224. """ R52+ViT-S/S16 w/ 2x2 patch hybrid @ 224 x 224.
""" """
backbone = _resnetv2((2, 4), **kwargs) backbone = _resnetv2((2, 4), **kwargs)
model_kwargs = dict( model_kwargs = dict(patch_size=2, embed_dim=384, depth=12, num_heads=6, **kwargs)
patch_size=2, embed_dim=384, depth=12, num_heads=6, hybrid_backbone=backbone, **kwargs) model = _create_vision_transformer_hybrid(
model = _create_vision_transformer('vit_small_r20_s16_p2_224', pretrained=pretrained, **model_kwargs) 'vit_small_r20_s16_p2_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
return model
@register_model
def vit_small_r20_s16_p2_384(pretrained=False, **kwargs):
""" R20+ViT-S/S16 w/ 2x2 Patch hybrid @ 384x384.
"""
backbone = _resnetv2((2, 4), **kwargs)
model_kwargs = dict(
embed_dim=384, patch_size=2, depth=12, num_heads=6, hybrid_backbone=backbone, **kwargs)
model = _create_vision_transformer('vit_small_r20_s16_p2_384', pretrained=pretrained, **model_kwargs)
return model return model
@ -197,18 +196,9 @@ def vit_small_r20_s16_224(pretrained=False, **kwargs):
""" R20+ViT-S/S16 hybrid. """ R20+ViT-S/S16 hybrid.
""" """
backbone = _resnetv2((2, 2, 2), **kwargs) backbone = _resnetv2((2, 2, 2), **kwargs)
model_kwargs = dict(embed_dim=384, depth=12, num_heads=6, hybrid_backbone=backbone, **kwargs) model_kwargs = dict(embed_dim=384, depth=12, num_heads=6, **kwargs)
model = _create_vision_transformer('vit_small_r20_s16_224', pretrained=pretrained, **model_kwargs) model = _create_vision_transformer_hybrid(
return model 'vit_small_r20_s16_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
@register_model
def vit_small_r20_s16_384(pretrained=False, **kwargs):
""" R20+ViT-S/S16 hybrid @ 384x384.
"""
backbone = _resnetv2((2, 2, 2), **kwargs)
model_kwargs = dict(embed_dim=384, depth=12, num_heads=6, hybrid_backbone=backbone, **kwargs)
model = _create_vision_transformer('vit_small_r20_s16_384', pretrained=pretrained, **model_kwargs)
return model return model
@ -217,18 +207,9 @@ def vit_small_r26_s32_224(pretrained=False, **kwargs):
""" R26+ViT-S/S32 hybrid. """ R26+ViT-S/S32 hybrid.
""" """
backbone = _resnetv2((2, 2, 2, 2), **kwargs) backbone = _resnetv2((2, 2, 2, 2), **kwargs)
model_kwargs = dict(embed_dim=384, depth=12, num_heads=6, hybrid_backbone=backbone, **kwargs) model_kwargs = dict(embed_dim=384, depth=12, num_heads=6, **kwargs)
model = _create_vision_transformer('vit_small_r26_s32_224', pretrained=pretrained, **model_kwargs) model = _create_vision_transformer_hybrid(
return model 'vit_small_r26_s32_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
@register_model
def vit_small_r26_s32_384(pretrained=False, **kwargs):
""" R26+ViT-S/S32 hybrid @ 384x384.
"""
backbone = _resnetv2((2, 2, 2, 2), **kwargs)
model_kwargs = dict(embed_dim=384, depth=12, num_heads=6, hybrid_backbone=backbone, **kwargs)
model = _create_vision_transformer('vit_small_r26_s32_384', pretrained=pretrained, **model_kwargs)
return model return model
@ -237,18 +218,9 @@ def vit_base_r20_s16_224(pretrained=False, **kwargs):
""" R20+ViT-B/S16 hybrid. """ R20+ViT-B/S16 hybrid.
""" """
backbone = _resnetv2((2, 2, 2), **kwargs) backbone = _resnetv2((2, 2, 2), **kwargs)
model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, hybrid_backbone=backbone, **kwargs) model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, **kwargs)
model = _create_vision_transformer('vit_base_r20_s16_224', pretrained=pretrained, **model_kwargs) model = _create_vision_transformer_hybrid(
return model 'vit_base_r20_s16_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
@register_model
def vit_base_r20_s16_384(pretrained=False, **kwargs):
""" R20+ViT-B/S16 hybrid.
"""
backbone = _resnetv2((2, 2, 2), **kwargs)
model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, hybrid_backbone=backbone, **kwargs)
model = _create_vision_transformer('vit_base_r20_s16_384', pretrained=pretrained, **model_kwargs)
return model return model
@ -257,18 +229,9 @@ def vit_base_r26_s32_224(pretrained=False, **kwargs):
""" R26+ViT-B/S32 hybrid. """ R26+ViT-B/S32 hybrid.
""" """
backbone = _resnetv2((2, 2, 2, 2), **kwargs) backbone = _resnetv2((2, 2, 2, 2), **kwargs)
model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, hybrid_backbone=backbone, **kwargs) model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, **kwargs)
model = _create_vision_transformer('vit_base_r26_s32_224', pretrained=pretrained, **model_kwargs) model = _create_vision_transformer_hybrid(
return model 'vit_base_r26_s32_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
@register_model
def vit_base_r26_s32_384(pretrained=False, **kwargs):
""" R26+ViT-B/S32 hybrid.
"""
backbone = _resnetv2((2, 2, 2, 2), **kwargs)
model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, hybrid_backbone=backbone, **kwargs)
model = _create_vision_transformer('vit_base_r26_s32_384', pretrained=pretrained, **model_kwargs)
return model return model
@ -277,8 +240,9 @@ def vit_base_r50_s16_224(pretrained=False, **kwargs):
""" R50+ViT-B/S16 hybrid from original paper (https://arxiv.org/abs/2010.11929). """ R50+ViT-B/S16 hybrid from original paper (https://arxiv.org/abs/2010.11929).
""" """
backbone = _resnetv2((3, 4, 9), **kwargs) backbone = _resnetv2((3, 4, 9), **kwargs)
model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, hybrid_backbone=backbone, **kwargs) model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, **kwargs)
model = _create_vision_transformer('vit_base_r50_s16_224', pretrained=pretrained, **model_kwargs) model = _create_vision_transformer_hybrid(
'vit_base_r50_s16_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
return model return model
@ -287,29 +251,9 @@ def vit_large_r50_s32_224(pretrained=False, **kwargs):
""" R50+ViT-L/S32 hybrid. """ R50+ViT-L/S32 hybrid.
""" """
backbone = _resnetv2((3, 4, 6, 3), **kwargs) backbone = _resnetv2((3, 4, 6, 3), **kwargs)
model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, hybrid_backbone=backbone, **kwargs) model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, **kwargs)
model = _create_vision_transformer('vit_large_r50_s32_224', pretrained=pretrained, **model_kwargs) model = _create_vision_transformer_hybrid(
return model 'vit_large_r50_s32_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
@register_model
def vit_large_r50_s32_224_in21k(pretrained=False, **kwargs):
""" R50+ViT-L/S32 hybrid.
"""
backbone = _resnetv2((3, 4, 6, 3), **kwargs)
model_kwargs = dict(
embed_dim=768, depth=12, num_heads=12, representation_size=768, hybrid_backbone=backbone, **kwargs)
model = _create_vision_transformer('vit_large_r50_s32_224_in21k', pretrained=pretrained, **model_kwargs)
return model
@register_model
def vit_large_r50_s32_384(pretrained=False, **kwargs):
""" R50+ViT-L/S32 hybrid.
"""
backbone = _resnetv2((3, 4, 6, 3), **kwargs)
model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, hybrid_backbone=backbone, **kwargs)
model = _create_vision_transformer('vit_large_r50_s32_384', pretrained=pretrained, **model_kwargs)
return model return model
@ -318,8 +262,9 @@ def vit_small_resnet26d_224(pretrained=False, **kwargs):
""" Custom ViT small hybrid w/ ResNet26D stride 32. No pretrained weights. """ Custom ViT small hybrid w/ ResNet26D stride 32. No pretrained weights.
""" """
backbone = resnet26d(pretrained=pretrained, in_chans=kwargs.get('in_chans', 3), features_only=True, out_indices=[4]) backbone = resnet26d(pretrained=pretrained, in_chans=kwargs.get('in_chans', 3), features_only=True, out_indices=[4])
model_kwargs = dict(embed_dim=768, depth=8, num_heads=8, mlp_ratio=3, hybrid_backbone=backbone, **kwargs) model_kwargs = dict(embed_dim=768, depth=8, num_heads=8, mlp_ratio=3, **kwargs)
model = _create_vision_transformer('vit_small_resnet26d_224', pretrained=pretrained, **model_kwargs) model = _create_vision_transformer_hybrid(
'vit_small_resnet26d_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
return model return model
@ -328,8 +273,9 @@ def vit_small_resnet50d_s16_224(pretrained=False, **kwargs):
""" Custom ViT small hybrid w/ ResNet50D 3-stages, stride 16. No pretrained weights. """ Custom ViT small hybrid w/ ResNet50D 3-stages, stride 16. No pretrained weights.
""" """
backbone = resnet50d(pretrained=pretrained, in_chans=kwargs.get('in_chans', 3), features_only=True, out_indices=[3]) backbone = resnet50d(pretrained=pretrained, in_chans=kwargs.get('in_chans', 3), features_only=True, out_indices=[3])
model_kwargs = dict(embed_dim=768, depth=8, num_heads=8, mlp_ratio=3, hybrid_backbone=backbone, **kwargs) model_kwargs = dict(embed_dim=768, depth=8, num_heads=8, mlp_ratio=3, **kwargs)
model = _create_vision_transformer('vit_small_resnet50d_s16_224', pretrained=pretrained, **model_kwargs) model = _create_vision_transformer_hybrid(
'vit_small_resnet50d_s16_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
return model return model
@ -338,8 +284,9 @@ def vit_base_resnet26d_224(pretrained=False, **kwargs):
""" Custom ViT base hybrid w/ ResNet26D stride 32. No pretrained weights. """ Custom ViT base hybrid w/ ResNet26D stride 32. No pretrained weights.
""" """
backbone = resnet26d(pretrained=pretrained, in_chans=kwargs.get('in_chans', 3), features_only=True, out_indices=[4]) backbone = resnet26d(pretrained=pretrained, in_chans=kwargs.get('in_chans', 3), features_only=True, out_indices=[4])
model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, hybrid_backbone=backbone, **kwargs) model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, **kwargs)
model = _create_vision_transformer('vit_base_resnet26d_224', pretrained=pretrained, **model_kwargs) model = _create_vision_transformer_hybrid(
'vit_base_resnet26d_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
return model return model
@ -348,6 +295,7 @@ def vit_base_resnet50d_224(pretrained=False, **kwargs):
""" Custom ViT base hybrid w/ ResNet50D stride 32. No pretrained weights. """ Custom ViT base hybrid w/ ResNet50D stride 32. No pretrained weights.
""" """
backbone = resnet50d(pretrained=pretrained, in_chans=kwargs.get('in_chans', 3), features_only=True, out_indices=[4]) backbone = resnet50d(pretrained=pretrained, in_chans=kwargs.get('in_chans', 3), features_only=True, out_indices=[4])
model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, hybrid_backbone=backbone, **kwargs) model_kwargs = dict(embed_dim=768, depth=12, num_heads=12, **kwargs)
model = _create_vision_transformer('vit_base_resnet50d_224', pretrained=pretrained, **model_kwargs) model = _create_vision_transformer_hybrid(
'vit_base_resnet50d_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
return model return model
Loading…
Cancel
Save