MareArts Computer Vision Study.: timm swin transformer v2 model review

5/20/2023

timm swin transformer v2 model review

refer to code:

import timm
import torch
from PIL import Image
from timm.data import resolve_data_config
from timm.data.transforms_factory import create_transform

# Load from Hub 🔥
model = timm.create_model(
    'hf-hub:nateraw/resnet50-oxford-iiit-pet',
    pretrained=True
)

# Set model to eval mode for inference
model.eval()

#resolve
print(resolve_data_config(model.pretrained_cfg, model=model))

# Get the labels from the model config
print(model.pretrained_cfg)

# labels
labels = model.pretrained_cfg['label_names']
print(labels)

top_k = min(len(labels), 5)

# Create Transform
transform = create_transform(**resolve_data_config(model.pretrained_cfg, model=model))

# Use your own image file here...
image = Image.open('boxer.jpg').convert('RGB')

# Process PIL image with transforms and add a batch dimension
x = transform(image).unsqueeze(0)

print(x.shape)
# Pass inputs to model forward function to get outputs
out = model(x)
print(out.shape)

# Apply softmax to get predicted probabilities for each class
probabilities = torch.nn.functional.softmax(out[0], dim=0)

print(probabilities)

# Grab the values and indices of top 5 predicted classes
values, indices = torch.topk(probabilities, top_k)

# Prepare a nice dict of top k predictions
predictions = [
    {"label": labels[i], "score": v.item()}
    for i, v in zip(indices, values)
]
print(predictions)

from torchsummary import summary
input_size = (3, 224, 224)
# Use torchsummary to print the model summary
summary(model, input_size)


# Remove the last layer
model = torch.nn.Sequential(*list(model.children())[:-1])
input_size = (3, 224, 224)
# Use torchsummary to print the model summary
summary(model, input_size)

First Summary:

----------------------------------------------------------------

Layer (type) Output Shape Param #

================================================================

Conv2d-1 [-1, 64, 112, 112] 9,408

BatchNorm2d-2 [-1, 64, 112, 112] 128

ReLU-3 [-1, 64, 112, 112] 0

MaxPool2d-4 [-1, 64, 56, 56] 0

Conv2d-5 [-1, 64, 56, 56] 4,096

BatchNorm2d-6 [-1, 64, 56, 56] 128

ReLU-7 [-1, 64, 56, 56] 0

Conv2d-8 [-1, 64, 56, 56] 36,864

BatchNorm2d-9 [-1, 64, 56, 56] 128

Identity-10 [-1, 64, 56, 56] 0

ReLU-11 [-1, 64, 56, 56] 0

Identity-12 [-1, 64, 56, 56] 0

Conv2d-13 [-1, 256, 56, 56] 16,384

BatchNorm2d-14 [-1, 256, 56, 56] 512

Conv2d-15 [-1, 256, 56, 56] 16,384

BatchNorm2d-16 [-1, 256, 56, 56] 512

ReLU-17 [-1, 256, 56, 56] 0

Bottleneck-18 [-1, 256, 56, 56] 0

Conv2d-19 [-1, 64, 56, 56] 16,384

BatchNorm2d-20 [-1, 64, 56, 56] 128

ReLU-21 [-1, 64, 56, 56] 0

Conv2d-22 [-1, 64, 56, 56] 36,864

BatchNorm2d-23 [-1, 64, 56, 56] 128

Identity-24 [-1, 64, 56, 56] 0

ReLU-25 [-1, 64, 56, 56] 0

Identity-26 [-1, 64, 56, 56] 0

Conv2d-27 [-1, 256, 56, 56] 16,384

BatchNorm2d-28 [-1, 256, 56, 56] 512

ReLU-29 [-1, 256, 56, 56] 0

Bottleneck-30 [-1, 256, 56, 56] 0

Conv2d-31 [-1, 64, 56, 56] 16,384

BatchNorm2d-32 [-1, 64, 56, 56] 128

ReLU-33 [-1, 64, 56, 56] 0

Conv2d-34 [-1, 64, 56, 56] 36,864

BatchNorm2d-35 [-1, 64, 56, 56] 128

Identity-36 [-1, 64, 56, 56] 0

ReLU-37 [-1, 64, 56, 56] 0

Identity-38 [-1, 64, 56, 56] 0

Conv2d-39 [-1, 256, 56, 56] 16,384

BatchNorm2d-40 [-1, 256, 56, 56] 512

ReLU-41 [-1, 256, 56, 56] 0

Bottleneck-42 [-1, 256, 56, 56] 0

Conv2d-43 [-1, 128, 56, 56] 32,768

BatchNorm2d-44 [-1, 128, 56, 56] 256

ReLU-45 [-1, 128, 56, 56] 0

Conv2d-46 [-1, 128, 28, 28] 147,456

BatchNorm2d-47 [-1, 128, 28, 28] 256

Identity-48 [-1, 128, 28, 28] 0

ReLU-49 [-1, 128, 28, 28] 0

Identity-50 [-1, 128, 28, 28] 0

Conv2d-51 [-1, 512, 28, 28] 65,536

BatchNorm2d-52 [-1, 512, 28, 28] 1,024

Conv2d-53 [-1, 512, 28, 28] 131,072

BatchNorm2d-54 [-1, 512, 28, 28] 1,024

ReLU-55 [-1, 512, 28, 28] 0

Bottleneck-56 [-1, 512, 28, 28] 0

Conv2d-57 [-1, 128, 28, 28] 65,536

BatchNorm2d-58 [-1, 128, 28, 28] 256

ReLU-59 [-1, 128, 28, 28] 0

Conv2d-60 [-1, 128, 28, 28] 147,456

BatchNorm2d-61 [-1, 128, 28, 28] 256

Identity-62 [-1, 128, 28, 28] 0

ReLU-63 [-1, 128, 28, 28] 0

Identity-64 [-1, 128, 28, 28] 0

Conv2d-65 [-1, 512, 28, 28] 65,536

BatchNorm2d-66 [-1, 512, 28, 28] 1,024

ReLU-67 [-1, 512, 28, 28] 0

Bottleneck-68 [-1, 512, 28, 28] 0

Conv2d-69 [-1, 128, 28, 28] 65,536

BatchNorm2d-70 [-1, 128, 28, 28] 256

ReLU-71 [-1, 128, 28, 28] 0

Conv2d-72 [-1, 128, 28, 28] 147,456

BatchNorm2d-73 [-1, 128, 28, 28] 256

Identity-74 [-1, 128, 28, 28] 0

ReLU-75 [-1, 128, 28, 28] 0

Identity-76 [-1, 128, 28, 28] 0

Conv2d-77 [-1, 512, 28, 28] 65,536

BatchNorm2d-78 [-1, 512, 28, 28] 1,024

ReLU-79 [-1, 512, 28, 28] 0

Bottleneck-80 [-1, 512, 28, 28] 0

Conv2d-81 [-1, 128, 28, 28] 65,536

BatchNorm2d-82 [-1, 128, 28, 28] 256

ReLU-83 [-1, 128, 28, 28] 0

Conv2d-84 [-1, 128, 28, 28] 147,456

BatchNorm2d-85 [-1, 128, 28, 28] 256

Identity-86 [-1, 128, 28, 28] 0

ReLU-87 [-1, 128, 28, 28] 0

Identity-88 [-1, 128, 28, 28] 0

Conv2d-89 [-1, 512, 28, 28] 65,536

BatchNorm2d-90 [-1, 512, 28, 28] 1,024

ReLU-91 [-1, 512, 28, 28] 0

Bottleneck-92 [-1, 512, 28, 28] 0

Conv2d-93 [-1, 256, 28, 28] 131,072

BatchNorm2d-94 [-1, 256, 28, 28] 512

ReLU-95 [-1, 256, 28, 28] 0

Conv2d-96 [-1, 256, 14, 14] 589,824

BatchNorm2d-97 [-1, 256, 14, 14] 512

Identity-98 [-1, 256, 14, 14] 0

ReLU-99 [-1, 256, 14, 14] 0

Identity-100 [-1, 256, 14, 14] 0

Conv2d-101 [-1, 1024, 14, 14] 262,144

BatchNorm2d-102 [-1, 1024, 14, 14] 2,048

Conv2d-103 [-1, 1024, 14, 14] 524,288

BatchNorm2d-104 [-1, 1024, 14, 14] 2,048

ReLU-105 [-1, 1024, 14, 14] 0

Bottleneck-106 [-1, 1024, 14, 14] 0

Conv2d-107 [-1, 256, 14, 14] 262,144

BatchNorm2d-108 [-1, 256, 14, 14] 512

ReLU-109 [-1, 256, 14, 14] 0

Conv2d-110 [-1, 256, 14, 14] 589,824

BatchNorm2d-111 [-1, 256, 14, 14] 512

Identity-112 [-1, 256, 14, 14] 0

ReLU-113 [-1, 256, 14, 14] 0

Identity-114 [-1, 256, 14, 14] 0

Conv2d-115 [-1, 1024, 14, 14] 262,144

BatchNorm2d-116 [-1, 1024, 14, 14] 2,048

ReLU-117 [-1, 1024, 14, 14] 0

Bottleneck-118 [-1, 1024, 14, 14] 0

Conv2d-119 [-1, 256, 14, 14] 262,144

BatchNorm2d-120 [-1, 256, 14, 14] 512

ReLU-121 [-1, 256, 14, 14] 0

Conv2d-122 [-1, 256, 14, 14] 589,824

BatchNorm2d-123 [-1, 256, 14, 14] 512

Identity-124 [-1, 256, 14, 14] 0

ReLU-125 [-1, 256, 14, 14] 0

Identity-126 [-1, 256, 14, 14] 0

Conv2d-127 [-1, 1024, 14, 14] 262,144

BatchNorm2d-128 [-1, 1024, 14, 14] 2,048

ReLU-129 [-1, 1024, 14, 14] 0

Bottleneck-130 [-1, 1024, 14, 14] 0

Conv2d-131 [-1, 256, 14, 14] 262,144

BatchNorm2d-132 [-1, 256, 14, 14] 512

ReLU-133 [-1, 256, 14, 14] 0

Conv2d-134 [-1, 256, 14, 14] 589,824

BatchNorm2d-135 [-1, 256, 14, 14] 512

Identity-136 [-1, 256, 14, 14] 0

ReLU-137 [-1, 256, 14, 14] 0

Identity-138 [-1, 256, 14, 14] 0

Conv2d-139 [-1, 1024, 14, 14] 262,144

BatchNorm2d-140 [-1, 1024, 14, 14] 2,048

ReLU-141 [-1, 1024, 14, 14] 0

Bottleneck-142 [-1, 1024, 14, 14] 0

Conv2d-143 [-1, 256, 14, 14] 262,144

BatchNorm2d-144 [-1, 256, 14, 14] 512

ReLU-145 [-1, 256, 14, 14] 0

Conv2d-146 [-1, 256, 14, 14] 589,824

BatchNorm2d-147 [-1, 256, 14, 14] 512

Identity-148 [-1, 256, 14, 14] 0

ReLU-149 [-1, 256, 14, 14] 0

Identity-150 [-1, 256, 14, 14] 0

Conv2d-151 [-1, 1024, 14, 14] 262,144

BatchNorm2d-152 [-1, 1024, 14, 14] 2,048

ReLU-153 [-1, 1024, 14, 14] 0

Bottleneck-154 [-1, 1024, 14, 14] 0

Conv2d-155 [-1, 256, 14, 14] 262,144

BatchNorm2d-156 [-1, 256, 14, 14] 512

ReLU-157 [-1, 256, 14, 14] 0

Conv2d-158 [-1, 256, 14, 14] 589,824

BatchNorm2d-159 [-1, 256, 14, 14] 512

Identity-160 [-1, 256, 14, 14] 0

ReLU-161 [-1, 256, 14, 14] 0

Identity-162 [-1, 256, 14, 14] 0

Conv2d-163 [-1, 1024, 14, 14] 262,144

BatchNorm2d-164 [-1, 1024, 14, 14] 2,048

ReLU-165 [-1, 1024, 14, 14] 0

Bottleneck-166 [-1, 1024, 14, 14] 0

Conv2d-167 [-1, 512, 14, 14] 524,288

BatchNorm2d-168 [-1, 512, 14, 14] 1,024

ReLU-169 [-1, 512, 14, 14] 0

Conv2d-170 [-1, 512, 7, 7] 2,359,296

BatchNorm2d-171 [-1, 512, 7, 7] 1,024

Identity-172 [-1, 512, 7, 7] 0

ReLU-173 [-1, 512, 7, 7] 0

Identity-174 [-1, 512, 7, 7] 0

Conv2d-175 [-1, 2048, 7, 7] 1,048,576

BatchNorm2d-176 [-1, 2048, 7, 7] 4,096

Conv2d-177 [-1, 2048, 7, 7] 2,097,152

BatchNorm2d-178 [-1, 2048, 7, 7] 4,096

ReLU-179 [-1, 2048, 7, 7] 0

Bottleneck-180 [-1, 2048, 7, 7] 0

Conv2d-181 [-1, 512, 7, 7] 1,048,576

BatchNorm2d-182 [-1, 512, 7, 7] 1,024

ReLU-183 [-1, 512, 7, 7] 0

Conv2d-184 [-1, 512, 7, 7] 2,359,296

BatchNorm2d-185 [-1, 512, 7, 7] 1,024

Identity-186 [-1, 512, 7, 7] 0

ReLU-187 [-1, 512, 7, 7] 0

Identity-188 [-1, 512, 7, 7] 0

Conv2d-189 [-1, 2048, 7, 7] 1,048,576

BatchNorm2d-190 [-1, 2048, 7, 7] 4,096

ReLU-191 [-1, 2048, 7, 7] 0

Bottleneck-192 [-1, 2048, 7, 7] 0

Conv2d-193 [-1, 512, 7, 7] 1,048,576

BatchNorm2d-194 [-1, 512, 7, 7] 1,024

ReLU-195 [-1, 512, 7, 7] 0

Conv2d-196 [-1, 512, 7, 7] 2,359,296

BatchNorm2d-197 [-1, 512, 7, 7] 1,024

Identity-198 [-1, 512, 7, 7] 0

ReLU-199 [-1, 512, 7, 7] 0

Identity-200 [-1, 512, 7, 7] 0

Conv2d-201 [-1, 2048, 7, 7] 1,048,576

BatchNorm2d-202 [-1, 2048, 7, 7] 4,096

ReLU-203 [-1, 2048, 7, 7] 0

Bottleneck-204 [-1, 2048, 7, 7] 0

AdaptiveAvgPool2d-205 [-1, 2048, 1, 1] 0

Flatten-206 [-1, 2048] 0

SelectAdaptivePool2d-207 [-1, 2048] 0

Linear-208 [-1, 37] 75,813

================================================================

Total params: 23,583,845

Trainable params: 23,583,845

Non-trainable params: 0

----------------------------------------------------------------

Input size (MB): 0.57

Forward/backward pass size (MB): 307.64

Params size (MB): 89.97

Estimated Total Size (MB): 398.18

Second Summary:

----------------------------------------------------------------

Layer (type) Output Shape Param #

================================================================

Conv2d-1 [-1, 64, 112, 112] 9,408

BatchNorm2d-2 [-1, 64, 112, 112] 128

ReLU-3 [-1, 64, 112, 112] 0

MaxPool2d-4 [-1, 64, 56, 56] 0

Conv2d-5 [-1, 64, 56, 56] 4,096

BatchNorm2d-6 [-1, 64, 56, 56] 128

ReLU-7 [-1, 64, 56, 56] 0

Conv2d-8 [-1, 64, 56, 56] 36,864

BatchNorm2d-9 [-1, 64, 56, 56] 128

Identity-10 [-1, 64, 56, 56] 0

ReLU-11 [-1, 64, 56, 56] 0

Identity-12 [-1, 64, 56, 56] 0

Conv2d-13 [-1, 256, 56, 56] 16,384

BatchNorm2d-14 [-1, 256, 56, 56] 512

Conv2d-15 [-1, 256, 56, 56] 16,384

BatchNorm2d-16 [-1, 256, 56, 56] 512

ReLU-17 [-1, 256, 56, 56] 0

Bottleneck-18 [-1, 256, 56, 56] 0

Conv2d-19 [-1, 64, 56, 56] 16,384

BatchNorm2d-20 [-1, 64, 56, 56] 128

ReLU-21 [-1, 64, 56, 56] 0

Conv2d-22 [-1, 64, 56, 56] 36,864

BatchNorm2d-23 [-1, 64, 56, 56] 128

Identity-24 [-1, 64, 56, 56] 0

ReLU-25 [-1, 64, 56, 56] 0

Identity-26 [-1, 64, 56, 56] 0

Conv2d-27 [-1, 256, 56, 56] 16,384

BatchNorm2d-28 [-1, 256, 56, 56] 512

ReLU-29 [-1, 256, 56, 56] 0

Bottleneck-30 [-1, 256, 56, 56] 0

Conv2d-31 [-1, 64, 56, 56] 16,384

BatchNorm2d-32 [-1, 64, 56, 56] 128

ReLU-33 [-1, 64, 56, 56] 0

Conv2d-34 [-1, 64, 56, 56] 36,864

BatchNorm2d-35 [-1, 64, 56, 56] 128

Identity-36 [-1, 64, 56, 56] 0

ReLU-37 [-1, 64, 56, 56] 0

Identity-38 [-1, 64, 56, 56] 0

Conv2d-39 [-1, 256, 56, 56] 16,384

BatchNorm2d-40 [-1, 256, 56, 56] 512

ReLU-41 [-1, 256, 56, 56] 0

Bottleneck-42 [-1, 256, 56, 56] 0

Conv2d-43 [-1, 128, 56, 56] 32,768

BatchNorm2d-44 [-1, 128, 56, 56] 256

ReLU-45 [-1, 128, 56, 56] 0

Conv2d-46 [-1, 128, 28, 28] 147,456

BatchNorm2d-47 [-1, 128, 28, 28] 256

Identity-48 [-1, 128, 28, 28] 0

ReLU-49 [-1, 128, 28, 28] 0

Identity-50 [-1, 128, 28, 28] 0

Conv2d-51 [-1, 512, 28, 28] 65,536

BatchNorm2d-52 [-1, 512, 28, 28] 1,024

Conv2d-53 [-1, 512, 28, 28] 131,072

BatchNorm2d-54 [-1, 512, 28, 28] 1,024

ReLU-55 [-1, 512, 28, 28] 0

Bottleneck-56 [-1, 512, 28, 28] 0

Conv2d-57 [-1, 128, 28, 28] 65,536

BatchNorm2d-58 [-1, 128, 28, 28] 256

ReLU-59 [-1, 128, 28, 28] 0

Conv2d-60 [-1, 128, 28, 28] 147,456

BatchNorm2d-61 [-1, 128, 28, 28] 256

Identity-62 [-1, 128, 28, 28] 0

ReLU-63 [-1, 128, 28, 28] 0

Identity-64 [-1, 128, 28, 28] 0

Conv2d-65 [-1, 512, 28, 28] 65,536

BatchNorm2d-66 [-1, 512, 28, 28] 1,024

ReLU-67 [-1, 512, 28, 28] 0

Bottleneck-68 [-1, 512, 28, 28] 0

Conv2d-69 [-1, 128, 28, 28] 65,536

BatchNorm2d-70 [-1, 128, 28, 28] 256

ReLU-71 [-1, 128, 28, 28] 0

Conv2d-72 [-1, 128, 28, 28] 147,456

BatchNorm2d-73 [-1, 128, 28, 28] 256

Identity-74 [-1, 128, 28, 28] 0

ReLU-75 [-1, 128, 28, 28] 0

Identity-76 [-1, 128, 28, 28] 0

Conv2d-77 [-1, 512, 28, 28] 65,536

BatchNorm2d-78 [-1, 512, 28, 28] 1,024

ReLU-79 [-1, 512, 28, 28] 0

Bottleneck-80 [-1, 512, 28, 28] 0

Conv2d-81 [-1, 128, 28, 28] 65,536

BatchNorm2d-82 [-1, 128, 28, 28] 256

ReLU-83 [-1, 128, 28, 28] 0

Conv2d-84 [-1, 128, 28, 28] 147,456

BatchNorm2d-85 [-1, 128, 28, 28] 256

Identity-86 [-1, 128, 28, 28] 0

ReLU-87 [-1, 128, 28, 28] 0

Identity-88 [-1, 128, 28, 28] 0

Conv2d-89 [-1, 512, 28, 28] 65,536

BatchNorm2d-90 [-1, 512, 28, 28] 1,024

ReLU-91 [-1, 512, 28, 28] 0

Bottleneck-92 [-1, 512, 28, 28] 0

Conv2d-93 [-1, 256, 28, 28] 131,072

BatchNorm2d-94 [-1, 256, 28, 28] 512

ReLU-95 [-1, 256, 28, 28] 0

Conv2d-96 [-1, 256, 14, 14] 589,824

BatchNorm2d-97 [-1, 256, 14, 14] 512

Identity-98 [-1, 256, 14, 14] 0

ReLU-99 [-1, 256, 14, 14] 0

Identity-100 [-1, 256, 14, 14] 0

Conv2d-101 [-1, 1024, 14, 14] 262,144

BatchNorm2d-102 [-1, 1024, 14, 14] 2,048

Conv2d-103 [-1, 1024, 14, 14] 524,288

BatchNorm2d-104 [-1, 1024, 14, 14] 2,048

ReLU-105 [-1, 1024, 14, 14] 0

Bottleneck-106 [-1, 1024, 14, 14] 0

Conv2d-107 [-1, 256, 14, 14] 262,144

BatchNorm2d-108 [-1, 256, 14, 14] 512

ReLU-109 [-1, 256, 14, 14] 0

Conv2d-110 [-1, 256, 14, 14] 589,824

BatchNorm2d-111 [-1, 256, 14, 14] 512

Identity-112 [-1, 256, 14, 14] 0

ReLU-113 [-1, 256, 14, 14] 0

Identity-114 [-1, 256, 14, 14] 0

Conv2d-115 [-1, 1024, 14, 14] 262,144

BatchNorm2d-116 [-1, 1024, 14, 14] 2,048

ReLU-117 [-1, 1024, 14, 14] 0

Bottleneck-118 [-1, 1024, 14, 14] 0

Conv2d-119 [-1, 256, 14, 14] 262,144

BatchNorm2d-120 [-1, 256, 14, 14] 512

ReLU-121 [-1, 256, 14, 14] 0

Conv2d-122 [-1, 256, 14, 14] 589,824

BatchNorm2d-123 [-1, 256, 14, 14] 512

Identity-124 [-1, 256, 14, 14] 0

ReLU-125 [-1, 256, 14, 14] 0

Identity-126 [-1, 256, 14, 14] 0

Conv2d-127 [-1, 1024, 14, 14] 262,144

BatchNorm2d-128 [-1, 1024, 14, 14] 2,048

ReLU-129 [-1, 1024, 14, 14] 0

Bottleneck-130 [-1, 1024, 14, 14] 0

Conv2d-131 [-1, 256, 14, 14] 262,144

BatchNorm2d-132 [-1, 256, 14, 14] 512

ReLU-133 [-1, 256, 14, 14] 0

Conv2d-134 [-1, 256, 14, 14] 589,824

BatchNorm2d-135 [-1, 256, 14, 14] 512

Identity-136 [-1, 256, 14, 14] 0

ReLU-137 [-1, 256, 14, 14] 0

Identity-138 [-1, 256, 14, 14] 0

Conv2d-139 [-1, 1024, 14, 14] 262,144

BatchNorm2d-140 [-1, 1024, 14, 14] 2,048

ReLU-141 [-1, 1024, 14, 14] 0

Bottleneck-142 [-1, 1024, 14, 14] 0

Conv2d-143 [-1, 256, 14, 14] 262,144

BatchNorm2d-144 [-1, 256, 14, 14] 512

ReLU-145 [-1, 256, 14, 14] 0

Conv2d-146 [-1, 256, 14, 14] 589,824

BatchNorm2d-147 [-1, 256, 14, 14] 512

Identity-148 [-1, 256, 14, 14] 0

ReLU-149 [-1, 256, 14, 14] 0

Identity-150 [-1, 256, 14, 14] 0

Conv2d-151 [-1, 1024, 14, 14] 262,144

BatchNorm2d-152 [-1, 1024, 14, 14] 2,048

ReLU-153 [-1, 1024, 14, 14] 0

Bottleneck-154 [-1, 1024, 14, 14] 0

Conv2d-155 [-1, 256, 14, 14] 262,144

BatchNorm2d-156 [-1, 256, 14, 14] 512

ReLU-157 [-1, 256, 14, 14] 0

Conv2d-158 [-1, 256, 14, 14] 589,824

BatchNorm2d-159 [-1, 256, 14, 14] 512

Identity-160 [-1, 256, 14, 14] 0

ReLU-161 [-1, 256, 14, 14] 0

Identity-162 [-1, 256, 14, 14] 0

Conv2d-163 [-1, 1024, 14, 14] 262,144

BatchNorm2d-164 [-1, 1024, 14, 14] 2,048

ReLU-165 [-1, 1024, 14, 14] 0

Bottleneck-166 [-1, 1024, 14, 14] 0

Conv2d-167 [-1, 512, 14, 14] 524,288

BatchNorm2d-168 [-1, 512, 14, 14] 1,024

ReLU-169 [-1, 512, 14, 14] 0

Conv2d-170 [-1, 512, 7, 7] 2,359,296

BatchNorm2d-171 [-1, 512, 7, 7] 1,024

Identity-172 [-1, 512, 7, 7] 0

ReLU-173 [-1, 512, 7, 7] 0

Identity-174 [-1, 512, 7, 7] 0

Conv2d-175 [-1, 2048, 7, 7] 1,048,576

BatchNorm2d-176 [-1, 2048, 7, 7] 4,096

Conv2d-177 [-1, 2048, 7, 7] 2,097,152

BatchNorm2d-178 [-1, 2048, 7, 7] 4,096

ReLU-179 [-1, 2048, 7, 7] 0

Bottleneck-180 [-1, 2048, 7, 7] 0

Conv2d-181 [-1, 512, 7, 7] 1,048,576

BatchNorm2d-182 [-1, 512, 7, 7] 1,024

ReLU-183 [-1, 512, 7, 7] 0

Conv2d-184 [-1, 512, 7, 7] 2,359,296

BatchNorm2d-185 [-1, 512, 7, 7] 1,024

Identity-186 [-1, 512, 7, 7] 0

ReLU-187 [-1, 512, 7, 7] 0

Identity-188 [-1, 512, 7, 7] 0

Conv2d-189 [-1, 2048, 7, 7] 1,048,576

BatchNorm2d-190 [-1, 2048, 7, 7] 4,096

ReLU-191 [-1, 2048, 7, 7] 0

Bottleneck-192 [-1, 2048, 7, 7] 0

Conv2d-193 [-1, 512, 7, 7] 1,048,576

BatchNorm2d-194 [-1, 512, 7, 7] 1,024

ReLU-195 [-1, 512, 7, 7] 0

Conv2d-196 [-1, 512, 7, 7] 2,359,296

BatchNorm2d-197 [-1, 512, 7, 7] 1,024

Identity-198 [-1, 512, 7, 7] 0

ReLU-199 [-1, 512, 7, 7] 0

Identity-200 [-1, 512, 7, 7] 0

Conv2d-201 [-1, 2048, 7, 7] 1,048,576

BatchNorm2d-202 [-1, 2048, 7, 7] 4,096

ReLU-203 [-1, 2048, 7, 7] 0

Bottleneck-204 [-1, 2048, 7, 7] 0

AdaptiveAvgPool2d-205 [-1, 2048, 1, 1] 0

Flatten-206 [-1, 2048] 0

SelectAdaptivePool2d-207 [-1, 2048] 0

================================================================

Total params: 23,508,032

Trainable params: 23,508,032

Non-trainable params: 0

----------------------------------------------------------------

Input size (MB): 0.57

Forward/backward pass size (MB): 307.64

Params size (MB): 89.68

Estimated Total Size (MB): 397.89

----------------------------------------------------------------

www.marearts.com

🙇🏻‍♂️

MareArts Computer Vision Study.

Pages

5/20/2023

timm swin transformer v2 model review

No comments:

Post a Comment

Labels

Archive