This repository contains a PyTorch implementation for the Plant Seedlings Classification problem. The goal is to classify images of plant seedlings into one of 12 different species based on their visual characteristics.

• Python 3.6 or later
• PyTorch
• scikit-learn
• matplotlib

1. Clone the repository:

git clone https://github.com/your-username/plant-seedlings-classification.git
cd plant-seedlings-classification

2. Download the dataset and place it in the data/train directory. The dataset should contain images of plant seedlings, organized into subdirectories corresponding to the different species.

3. Run the training script with the desired configuration:

python train.py --model resnet50 --data_dir data/train --batch_size 64 --num_epochs 50 --learning_rate 0.001

The available options are:

• --model: The model architecture to use (squeezenet, resnet50, mobilenetv2, efficientnetb0, inceptionv3).
• --data_dir: The directory containing the dataset.
• --batch_size: The batch size for training and evaluation.
• --num_epochs: The number of training epochs.
• --learning_rate: The learning rate for the optimizer.
• --num_classes: The number of classes in the dataset (default: 12).
• --test_size: The fraction of the dataset to be used as the test set (default: 0.2).
• --val_size: The fraction of the dataset to be used as the validation set (default: 0.2).

4. The training process will start, and the loss, accuracy, precision, recall, and F1-score for both the training and validation sets will be logged and displayed in the console.

5. After training, the model weights will be saved in the results directory, along with plots showing the loss, accuracy, precision, recall, and F1-score over the training epochs.

• train.py: The main script for training the model.
• load_dataset.py: Contains functions for loading and splitting the dataset.
• net.py: Defines the model architectures (SqueezeNet, ResNet50, MobileNetV2, EfficientNetB0, InceptionV3).

This project is licensed under the MIT License.

• The dataset used in this project is the Plant Seedlings Classification dataset from Kaggle.