README.md

---
license: mit
pipeline_tag: image-classification
tags:
- image-classification
- pytorch
- appliances
datasets:
- GBooster/imclass-dataset
---

# ImClass Model

Классификатор изображений бытовой техники (5 классов):
laptop, refrigerator, smartphone, tv, washing_machine

## Использование

```python
from huggingface_hub import PyTorchModelHubMixin
import torch.nn as nn
from torchvision import models as tvm

class ImClassModel(nn.Module, PyTorchModelHubMixin):
    def __init__(self, num_classes=5):
        super().__init__()
        base = tvm.resnet18(weights=None)
        base.fc = nn.Linear(base.fc.in_features, num_classes)
        self.model = base
    def forward(self, x):
        return self.model(x)

model = ImClassModel.from_pretrained("GBooster/imclass-model")
model.eval()
```

## Обучающие данные

Модель обучена на [GBooster/imclass-dataset](https://huggingface.co/datasets/GBooster/imclass-dataset).