---
license: apache-2.0
language:
  - ja
library_name: pytorch
tags:
  - recommendation
  - tabular
  - ft-transformer
model_name: gohan-product-recommendation
repo_type: model
---

# Gohan Product Recommendation Model

This model provides product recommendations for Gohan (rice) products using FT-Transformer architecture.

## Model Information
- **Architecture**: FT-Transformer
- **Training Epochs**: 30
- **Validation Performance**: 0.7736
- **Input Features**: Categorical features
- **Output**: Product recommendations

## Usage

```python
from inference import load_model, predict

# Load the model
model = load_model()

# Make predictions
predictions = predict(model, input_data)
```

## Files
- `epoch_030_p30_0.7736.pt`: Trained PyTorch model (Git LFS)
- `gohan_product_master_data.csv`: Product master data
- `encoders/`: JSON-encoded feature encoders
- `configs/config.json`: Model configuration

## License
Apache-2.0