Custom Transformer Model for MEXC Price Prediction

中文版

Model Description

Model Architecture

• Input Dimension: 13
• Model Dimension: 64
• Number of Heads: 8
• Number of Layers: 2
• Output Dimension: 1

Training Data

Training Details

• Optimizer: Adam
• Learning Rate: 0.001
• Loss Function: Mean Squared Error (MSE)
• Batch Size: 32
• Number of Epochs: 50

Usage

1. Load the model and configuration:
   python

   1import torch 2import torch.nn as nn 3from transformers import AutoConfig 4 5class CustomTransformerModel(nn.Module): 6 def __init__(self, config): 7 super(CustomTransformerModel, self).__init__() 8 self.embedding = nn.Linear(config.input_dim, config.model_dim) 9 self.encoder_layer = nn.TransformerEncoderLayer(d_model=config.model_dim, nhead=config.num_heads, batch_first=True) 10 self.transformer_encoder = nn.TransformerEncoder(self.encoder_layer, num_layers=config.num_layers) 11 self.fc = nn.Linear(config.model_dim, config.output_dim) 12 13 def forward(self, src): 14 src = self.embedding(src) 15 output = self.transformer_encoder(src) 16 output = self.fc(output[:, -1, :]) 17 return output 18 19config = AutoConfig.from_pretrained("your-username/mexc_price_model", config_file_name="BTC_USDT.json") 20model = CustomTransformerModel(config) 21model.load_state_dict(torch.load("model_repo/mexc_price.pth")) 22model.eval()
2. Prepare input data and make predictions:
   python

   1import numpy as np 2from sklearn.preprocessing import StandardScaler 3 4new_data = np.array([ 5 [1.727087e+09, 63483.9, 63426.2, 63483.9, 63411.6, 1193897.0, 7.575486e+06, 63483.8, 63426.2, 63483.9, 63411.6, 0.00, 0.0, 0.0] 6]) 7 8scaler = StandardScaler() 9new_data_scaled = scaler.fit_transform(new_data) 10input_tensor = torch.tensor(new_data_scaled, dtype=torch.float32).unsqueeze(1) 11 12with torch.no_grad(): 13 prediction = model(input_tensor) 14 15predicted_value = prediction.squeeze().item() 16print(f"Predicted Value: {predicted_value}")

License