| --- |
| license: mit |
| datasets: |
| - scikit-learn/iris |
| language: |
| - en |
| base_model: |
| - NeuralNine999/INET |
| pipeline_tag: tabular-classification |
| tags: |
| - biology |
| --- |
| |
| # INet - PyTorch Iris Classifier |
|
|
| ## Overview |
| INet is a simple fully-connected neural network trained on the Iris dataset using PyTorch. |
| It classifies iris flowers into 4 categories based on 4 features: sepal length, sepal width, petal length, and petal width. |
|
|
| ## Model Architecture |
| - Input: 4 features |
| - Hidden layers: 64 → 32 → 16 → 8 neurons (ReLU activations) |
| - Output: 4 classes |
|
|
| Architecture flow: |
| Input(4) → Linear(64) → ReLU → Linear(32) → ReLU → Linear(16) → ReLU → Linear(8) → ReLU → Linear(4) |
|
|
| - Loss: CrossEntropyLoss |
| - Optimizer: Adam, lr=0.01 |
| - Epochs: 30 |
|
|
| ## Files |
| - inet.pth → Trained model weights |
| - model.py → Contains INet class and architecture |
| - README.md → This file |
|
|
| ## How to Load |
|
|
| ```python |
| import torch |
| from model import INet # make sure INet class is in model.py |
| |
| model = INet() |
| model.load_state_dict(torch.load("inet.pth")) |
| model.eval() |
| |
| # Example usage: |
| sample_input = torch.tensor([[5.1, 3.5, 1.4, 0.2]]) |
| pred = model(sample_input) |
| pred_class = pred.argmax(dim=1).item() |
| print(pred_class) |
| ``` |
|
|
| ## Notes |
|
|
| * Make sure PyTorch is installed correctly |
| |
| ```python |
| pip install torch |
| ``` |
|
|
| * The model expects input as a tensor of shape [batch_size, 4] with float32 values. |
