---
tags:
- model_hub_mixin
- pytorch_model_hub_mixin
license: mit
---

# LSTMBitCoin: Bitcoin Price Prediction with LSTMs

## Model Details

- **Model Architecture:** Long Short-Term Memory (LSTM) Network  
- **Framework:** PyTorch  
- **Input Shape:** Time series sequences of Bitcoin price data (single feature)  
- **Output:** Predicted Bitcoin price for the next timestep  
- **Dataset:** [Bitcoin Historical Data](https://www.kaggle.com/datasets/mczielinski/bitcoin-historical-data)

## Model Description

The **LSTMBitCoin** model is a **recurrent neural network (RNN)** designed for Bitcoin price prediction. It utilizes a **single-layer LSTM** with **64 hidden units** followed by a **fully connected feedforward network**. The architecture captures temporal dependencies in historical Bitcoin price data.

## Training Details

- **Optimizer:** Adam  
- **Batch Size:** 64  
- **Loss Function:** Mean Squared Error (MSE)  
- **Number of Epochs:** 10  
- **Dropout:** 50%  
- **Activation Functions:** ReLU in the feedforward layers  

### Model Architecture

```python
class LSTMBitCoin(nn.Module, PyTorchModelHubMixin):
    def __init__(self):
        super(LSTMBitCoin, self).__init__()
        self.lstm = nn.LSTM(
            input_size=1, hidden_size=64, num_layers=1, batch_first=True, dropout=0.5
        )
        self.seq1 = nn.Sequential(
            nn.Flatten(),
            nn.ReLU(),
            nn.Dropout(0.5),
            nn.Linear(1280, 32),
            nn.ReLU(),
            nn.Dropout(0.5),
            nn.Linear(32, 1),
        )

    def forward(self, x):
        x, _ = self.lstm(x)
        x = self.seq1(x)
        return x.flatten()

This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration:
- Library: [More Information Needed]
- Docs: [More Information Needed]