resnet18-cifar10

#Generated using ChatGPT

---

tags:

• image-classification
• vision
• pytorch
• resnet
• cifar10
• transfer-learning license: mit library_name: timm

---

ResNet18 Fine-Tuned on CIFAR-10

Model Overview

This is a fine-tuned ResNet18 model for classifying images into 10 categories from the CIFAR-10 dataset. It was trained using transfer learning on a subset of ImageNet-pretrained weights and optimized for small image classification tasks.

Model Details

• Model Architecture: ResNet18
• Pre-trained on: ImageNet
• Fine-tuned on: CIFAR-10 dataset
• Number of Classes: 10
• Input Image Size: 32x32 (resized to 224x224)
• Framework: PyTorch

Intended Use

This model is intended for:

• Image classification tasks related to CIFAR-10 categories.
• Transfer learning demonstration for students and researchers.
• Educational purposes in computer vision.

Training Data

• Dataset: CIFAR-10
• Classes:
  • Airplane
  • Automobile
  • Bird
  • Cat
  • Deer
  • Dog
  • Frog
  • Horse
  • Ship
  • Truck
• Train/Test Split: 50,000 training images, 10,000 test images.

Training Procedure

• Preprocessing:
  • Images resized to 224x224
  • Normalized with mean [0.5, 0.5, 0.5] and std [0.5, 0.5, 0.5]
• Loss Function: CrossEntropyLoss
• Optimizer: Adam (learning rate: 0.001)
• Epochs: 5
• Batch Size: 32
• Hardware Used: Google Colab GPU (Tesla T4)

Performance

How to Use

You can load the model using torch.hub:

import torch
import torchvision.models as models
import torch.nn as nn

# Load fine-tuned ResNet18 model
model = models.resnet18(pretrained=False)
num_ftrs = model.fc.in_features
model.fc = nn.Linear(num_ftrs, 10)

# Load saved model weights
model.load_state_dict(torch.hub.load_state_dict_from_url(
    'https://huggingface.co/your_username/resnet18-cifar10/resolve/main/resnet18_cifar10.pth',
    map_location=torch.device("cpu")
))
model.eval()